October 2008

Mal's Musings

Malcolm A Traill

Dated 27/10/2008






The use of Lithium in the treatment of an elderly patient (LC) with lung cancer was presented in an earlier Chapter[1]. The key changes noted were an impressive reduction in tumour size, an elevation of the peripheral blood NK cell count and possibly a small reduction in the serum tumour marker, Carcinoembryonic antigen (CEA). There were important features absent, such as an histological diagnosis and adequate follow-up but, there seemed enough points to warrant further investigations into the Lithium treatment with an appropriate, cooperative patient. This report is a careful analysis of the features of interest observed when a patient (PC) with prostate cancer was treated with Lithium up until his radical prostatectomy. The main features of interest are a probable size reduction of both primary tumour and the prostate gland, a rise in the peripheral blood NK cell count (probably dependent upon concurrent “non-self” stimuli), cytological changes consistent with patchy apoptosis, that not well demonstrated until viewing the prostatectomy specimen some two months after the commencement of the Lithium treatment. The two cancer patients shall be compared and contrasted.

Basic comparison

Organ site                                     Cell type                                      Doubling Time (measure)

Lung Cancer (LC)

Adenocarcinoma (?), CEA+

~101 day (X-ray change)

Prostate (PC)

Adenocarcinoma, PSA+

~4.75 year (PSA rise)


The patient: He was 68 years old, with a known rising level of Prostate Specific Antigen (PSA) for some years handled conservatively. His brother had carcinoma of the prostate diagnosed several years previously, and his mother’s family had a sister and a mother with probable gastro-intestinal cancers. When the PSA value reached 12.7 μg/L, he decided to submit to the diagnostic work-up indicated as preparation for radical prostatectomy. With the basic diagnosis of cancer presumed, and with full knowledge of the aims and likely risks and benefits of the Lithium treatment, he had baseline tests in order to assess the subsequent Lithium treatment, and then proceeded to have the Lithium doses, whilst awaiting the other standard investigations. At no stage did the Lithium treatment interfere with or delay the progress to radical prostatectomy. The treatment was considered a possible method of reducing metastasis before or during surgery. Following clear a CT body scan and bone scan, the subsequent surgical specimen was reported to show Gleason levels 3+4 = 7 and pelvic lymph nodes were clear. Six weeks post-operatively, the PSA was reported to be 0.01 μg/L, not clearly differing from “black-box” noise. His weight was 66 kg, height 173 cm and Body Mass Index was 22 kg.m-2.

Lithium Treatment: Throughout and until the present, Lithium carbonate was administered as two bolus doses, each of 500 mg, as tablets broken into four pieces, taken six hour apart on one day (almost always at 12:00 h and at 18:00 h) followed by a day without any doses; the next paired boluses to be taken the following day, being 42 h after the previous second bolus (ie an alternate day regime). This was based upon the observations in animals[2], and may not be optimal but, since there is no other reference regime, it has been used and adhered-to closely, with estimated 6 h +/- 5 min between the boluses on the treatment days. For simplicity, the two 500 mg boluses on the one day shall be referred-to collectively as Rx. Sample serum level Lithium values are shown in the Figure 1. The levels are probably lower than typical, and may be attributed to an individual variation[3], which may not be explained yet. Fluid input was consciously maintained plentiful. Other aspects of diet and lifestyle were, and remained essentially unchanged except for the inpatient interval: usually having one multivitamin tablet, a single 1g fish oil capsule and ~100 mg aspirin per day, with meat and eggs, vegetables and fruit, ~4 mugs of coffee per day and only occasional alcohol.



Figure 1: The graph shows the levels of serum lithium obtained on various days after weeks in the regime, as relating to Rx. The laboratory reported the pre-Rx level as <0.4 mmol/L, so a value of 0.3 mmol/L has been assumed. The times of the individual bolus doses of 500 mg are shown as dashed blue lines. The values are lower than would normally be expected (probably individual variation). The lines are dashed, because the point-to-point changes would not be simple and linear. At no stage was a “toxic” level of Lithium attained and there was no appreciable accumulation over time.

Lithium Rx and days


                Rx day  Investigations                                   Procedure

0 (start)

PSA, Creatinine

Lymphocyte surface markers

CT of lower abdomen and pelvis



Lymphocyte surface markers






CT of lower abdomen & pelvis

Bone scan



PSA, Creatinine

Lymphocyte surface markers




Radical prostatectomy


Lymphocyte surface markers




Effects of Lithium Rx

Rx and Renal Function: PC noticed that he seemed to have polyuria on the days with Rx. Accordingly, urine output was measured:


                                                Rx day            No Rx

Days measured                        11                    11

Urine output – Median             5.1 L                3.5 L

                        Mean               5.16 L              3.64 L

Standard deviation                    0.47 L              0.45 L

Variance                                   0.2                   0.19


This indicated that there may be an increased urine output on Rx days, but there were uncontrolled factors.


Accordingly, PC collected hourly urine output volumes during waking hours, with an attempt to match the fluid inputs on the two days. This showed that there appeared to be a diuresis particularly during the second - fourth hour following each bolus dose.





FIGURE 2: Urine was collected hourly during waking hours, with an attempt to match the fluid input on the first day with the same input on the second day. Fluid in food was disregarded, so that the results are, at best, approximate.




Figure 3: The results for Figure 2 presented as hourly output differences, showing a relative dehydration before the Rx, then a diuresis during each 2nd to 4th+ hour following each of the 500 mg Lithium carbonate bolus doses.


The delay, by an hour, for the diuresis following each bolus dose of Lithium carbonate is consistent with a metabolic process, rather than a simple osmotic effect. Whilst Lithium has been known for a long time to cause a diuresis, there may not be awareness that it can happen so early after bolus dosing. Presumably it reflects an intracellular Lithium level. Since the first bolus dose for the Rx produced a (probable) peak Lithium serum level of only ~0.6 mmol/L, the renal response seems sensitive.

The Rx produced no side effects (other than the polyuria), the serum creatinine remained steady, no antinuclear antibodies or anti-ds-DNA antibodies developed and with the blood specimen taken at the measured Lithium peak (0.7 mmol/L) the serum corrected calcium was 2.27 mmol/L (Reference Range 2.10-2.60) and inorganic phosphate 1.2 mmol/L (RR 0.6-1.4).


Prostate Gland & Tumour Size: PC had Computerized tomography (CT) on 21st of May and 17th of June. These were performed at different sites, with differing criteria, such as the distance between “slices” on the first occasion being 8 mm, on the second, 5 mm. There was also uncertainty in assessing how the “slices” by both procedures related. The ultimate assessment was on the prostatectomy specimen. The gross specimen’s dimensions (13th of August) were recorded by the examining pathologist, and subsequently, 2 transverse slices of the tissue were recovered from the laboratory. One would appear to have been sectioned close to the largest dimension of the tumour (and gland) and is presented in Figure 4.

                                                                                                                   FIGURE 4

{C}{C}{C}{C}{C}{C}{C}{C}{C}{C}{C}{C}Text Box: Tumour

Text Box: Right posterior


Figure 4: Cross section of the prostatectomy specimen, the section probably close to the largest dimension of the tumour. The specimen is orientated as it is usually presented in CT scans, with the right posterior part of the gland at the bottom left of the Figure – as though viewed from the feet. The grey tumour (cancer) is located in the right posterolateral angle of the gland and, as assessed macroscopically, is about 1.2-3 cm in diameter, with irregular boundaries. The yellow area anterior to, or in the grey tumour, is probably produced by intracellular lipid, and is of uncertain significance. The transverse distance (blue) is shown, and, depending upon exact reading, may pass through tumour, although, in reading from the CT scan images, a non-tumour dimension would be read, probably.  The “right oblique,” (pink) passes through the tumour.  The “left oblique,” (green) does not pass through tumour, and can be regarded as a reference length including the left postero-lateral lobe.  The upper 2/3 of the gland shows benign enlargement changes, with adenomatous and cyst formation.


Gross Prostate Gland Size Changes: This is shown in Figure 5, and compared with the observed changes in the tumour size of patient LC.




Figure 5: (Note the semi-logarithmic format.) Throughout, the volumes are calculated from the basic equation 4/3πr3, with individual axes used for each r. With a non-spherical object such as the prostate, the result can only be an approximation to be used for comparison purposes.  The initial CT axes were provided by the Radiologist without knowledge of the use to which they may be put. The gross specimen axes were provided by the pathologist, again without knowledge of potential use. The middle values were read by the author. In the case of LC, two axes were read from X-ray films, the third axis assumed to be the mean of the first two. The vertical scale on the Figure is logarithmic.


The degree of this result was surprising and some support from relative measurements was sought, and shown in Figure 6.



Figure 6: The length of the “right oblique” distance (see Figure 4) as a ratio to the width and the “left oblique.” There appears to be a relative reduction in the “right oblique” (which includes tumour) and the width which, as viewed by imaging, would probably reflect largely infiltrating tumour-free gland anterior to the tumour. The ratios reflect the gross gland volume changes. The ratio with the “left oblique” showed an initial reduction, then an appreciable rise. The former may indicate a relative decrease in the tumour size on the right, but the latter is more difficult to explain, unless there has been a relative reduction in the size of the left postero-lateral part of the gland. (Refer to Figure 7 and text.)



Figure 7: This representation indicates that the “left oblique” was relatively large in relation to the width initially, to fall steadily throughout. Against the “right oblique” (with tumour), it was small initially, to become only slightly smaller than the right oblique, then smaller again, to match the width. See text.)


Blood Lymphocytes: Patient LC demonstrated peripheral blood lymphocyte changes, as revealed by assessing the surface markers.



Figure 8: This composite graph shows the lymphocyte surface makers of PC in bold lines, and the equivalent markers for LC using the same colours as for PC, but with dotted lines.  A single estimation of PC’s Regulatory T cells (Treg) is shown, expressed as an absolute count. The graphs are superimposed as though the treatments started together, with the dates referring to PC’s.


Patient PC showed an initial fall in CD4 T cells, with a later rise of both CD4 and CD8 T cells and Natural Killer (NK) cells, the latter close to doubling.


            Patient            Earliest NK    Peak NK         Increase (%)

            LC                   210x109/L        290 x109/L       +138%

            PC                   213 x109/L       353 x109/L       +166%


Post operatively, all fell; the dynamics of this unknown. The rises may be attributed to the effects of Lithium – in which case, it may be one of the few, if not the only, agents that can stimulate peripheral blood NK cell levels (&/or reduce consumption). Post-operatively, the NK, CD4 and CD8 cell numbers fell. The reason for the fall is unclear but, in the case of the NK cells, the peripheral blood initial rise in NK cell number in response to Lithium Rx (greater than for CD8 T cells) may require concurrent Lithium Rx andnon-self” stimuli provided by the cancer – when one stimulus type is removed, the NK cell response falls away. The CD8 T cells that reflect the Cytotoxic T cells (CTL) show a similar fall post-operatively. Influences, other than the tumour/gland removal could be operative stress and histamine release following morphine, although, with an epidural anaesthetic, this would not be great. Both these influences would be expected to be transitory and dissipated by the time of the next blood test.

In comparing these results with those of patient LC (remember that she had a much larger and more rapidly growing tumour, which showed an impressive decrease in volume with Rx), the fall in the CD4 and CD8 T cells may have been associated with the decreasing tumour size, as by apoptosis with T cell consumption, loss of stimulating epitopes on the tumour cells or toxic effects. LC’s lymphocyte decreases would seem to become apparent about 34 days into the Rx (about ⅓ the doubling time).

B cells for both patients showed little change.

Regulatory T cells (Treg) were counted as a percentage of CD4 T cells (see Figure for timing). Sadly a pre-Rx measure was not performed - but better late than never ! This non-standardized test expressed the Treg as 7.92% of total peripheral CD4 cells, with an “expected” range of 5-10% of CD4 T cells. Since PC had no known condition or therapy that could be associated with elevation or suppression of the Treg numbers, the assumption may be that the Treg numbers were in the “normal” range pre-Rx. In which case, Lithium Rx is unlikely to cause appreciable suppression or elevation of Treg numbers in the peripheral blood.

Serum Prostate Specific Antigen levels




Figure 9: The PSA levels (blue) have been expressed as log10 against time in months. Regression lines (least squares) are presented, the pink line excluding the last 3 points, the brown and longer line, excluding the last two points. Both regression lines are almost identical. The third last point corresponds to PSA=12.7 μg/L. The PSA of 12.7 μg/L falls on, or very close, to both regression lines, meaning that the PSA value should be, or very close to, the “true” PSA level. The final two points have the laboratories’ claimed Standard Deviations for day-to-day variations shown. There might be a trend downwards with the start of Lithium Rx, and there might be some trend upwards at the time of prostatectomy – to be related to the Histology.  (See text.)   


Photomicroscopy and Enlargement

Photomicrographs will be presented. All have been taken with the same equipment: a microscope with objectives x4, x10, x40, and x100 (oil). There was an “ocular” lens of x5, leading to a film negative 24x36 mm. If an image taken under the objective x40 is viewed with its subsequent greater dimension on a screen ~23 cm wide, the magnification will be:

            40 x 5 x (23/3.6) = 1,278.

Accordingly, only the objectives used will be stated.


Histopathology: Recut sections of PC’s biopsy and the excision specimen were obtained and were reviewed using Haematoxylin and Eosin staining only, which was limiting. The cancer was multifocal, being found at the right anterior and posterior upper, mid and lower regions and left lower anterior, with extensive prostate intraepithelial neoplasia (PIN) – there can be an assumed total gland pre-cancerous, epithelial and genetic defect.

Apart from the obvious presence of cancer, PIN and benign changes, there are some features of particular interest.

Comment will start with the gross specimen:

a)      Degeneration. There are numerous areas of carcinomatous glandular epithelial “degeneration,” a non-specific term that will be used in the absence of more specific characterization of the process. This can be seen under low power, because of the darkened filling of the stromal “tubes” retaining the degenerating glandular cells that previously lined the “tubes.” (see Figure 10).


{C}{C}{C}{C}FIGURE 10



{C}{C}{C}{C}{C}{C}{C}{C}{C}{C}Figure 10: Right posterior lobe. The surgical excision line is at the top, marked with a blue dye. To the mid left (green arrow) there is a moderately differentiated malignant gland, showing some early, patchy degeneration of its lining cells.   Further left are metastatic adenocarcinoma glands in a perineural lymphatic (blue) beneath the gland’s capsule. The cells show early degeneration. Through the right half there are scattered dark areas made conspicuous by the pyknotic nuclei of the degenerating glandular cells, the compaction of the nuclei and the loss of gland lumina (only some are shown with black arrows). (Objective x4)


b)      Under higher power, the glandular cells seem to become more eosinophilic, lose attachment, and develop pyknotic and irregular nuclei, consistent with apoptosis.





Figure 11: Degenerating cancer glands are scattered through the field. Noticeable are the pyknotic nuclei, eosinophilic cytoplasm, lack of cell polarity and cohesion and loss of gland lumina. Inflammatory cells are sparse in the stroma, although their presence amongst the degenerating cells remains a possibility. (Objective x10)



{C}{C}{C}{C}{C}{C}{C}{C}{C}{C} {C}{C}

Figure 12: Cancer deposit in a perineural lymphatic. The nerve is mid-lower, with an arteriole above it to the right. The degenerative adenocarcinoma cells, with pyknotic nuclei, nuclear crowding, and loss of gland lumen, are shown mid field. At the lower right are clumps of what appear to be bare nuclear debris (blue arrow) that are within the lymphatic channel, as though collected there by drifting, sedimentation and silting. Some small cells in the stroma appear to be inflammatory cells (red arrows). They are generally remote from the degeneration in the lymphatic channel. There are some spilt red blood cells in the lymphatic channel, possibly from surgical trauma. (Objective x40)


c)      The hallmark of the carcinomatous glandular epithelial degeneration seems to be loss of their lumina.

d)      Inflammatory cells are sparse in the cancer areas, and possibly more conspicuous in the more benign areas. This means that effector cells (if any) are retained for only very brief intervals in the stroma, before entering the epithelial layer and becoming lost in the degeneration. Assessment of this would require other techniques. The degeneration also does not appear to stimulate any appreciable (if at all) inflammatory reactive process. Such a conclusion would be consistent with the degeneration being basically an early apoptosis process.

e)      The degeneration does not appear to be related to artefacts created by either a delay in formaldehyde fixation or hypoxia, the effects of both usually seen inside large tumour masses removed in routine surgical practice. Photomicrographs were generally taken from areas close to the excision margin, in order to protect against the possibility of delayed fixation artefacts. Features that make delayed fixation unlikely are :

                                               i            Pronounced degenerative changes noted near the excision line


                                                                                            FIGURE 13


Figure 13: Adenocarcinoma cells in a perineural lymphatic near the excision line (right upper). There are cancer cells glands at lower left with mild degeneration. About mid-field there are degenerative cells and debris, with a less affected part of the gland to the left and beneath. What appear to be masses of nuclear debris (black) are scattered seemingly within the lymphatic lumen upper left, mid and lower right. Such deposits are relatively rare, and may represent drifting and sedimenting collections from degeneration “upstream” in the lymphatic channel. Inflammatory cells are sparse. (Objective x10)



Figure 14: This is an enlargement from the previous photomicrograph, to demonstrate the degeneration of the glands, the loss of cellular cohesion, loss of polarity and loss of the lumen, to produced unattached carcinoma cells, some still with apparently unaffected nuclei. The juxtaposition of degenerating cells against organized glandular cells is shown (eg lower right). Note the irregular, pyknotic nuclei (lower left to mid field) and the mass of presumed nuclear debris (upper left), below which there may be some early degenerative cells. Near the centre there appears to be a smaller nuclear debris deposit.  A mucoid fluid filled some spaces. (Objective x40) 

                                             ii            The changes are scattered through the tissue (Figures 10 & 11)

                                            iii            Progression from apparently near-vital cells to degeneration can be observed (Figures 13, 14, 15 & 16)



Figure 15: A later stage in degeneration. Cells at lower, right of centre appear to have been recently dislodged from the gland basement membrane. Note the irregular shapes of the nuclei generally, particularly the angular forms. There is cytoplasmic debris. A few inflammatory cells may be present in the stroma. (Objective x40)




Figure 16: Left of centre is a tuft of columnar epithelial cells without marked degenerative changes, unlike those on the opposite of the gland lumen. Lower down, the degenerative epithelial cells show transition to more viable cells.  To the right, cells have dislodged, with contraction and obliteration of the lumen. Inflammatory cells in the stroma are not conspicuous. (Objective x40)





{C}Figure 17: Carcinoma glands showing patchy features that may be early degenerative changes (black arrows). Lower, left of centre, are what appear to be clumps of pyknotic nuclei. (Objective x40)

                                           iv            In some areas nuclear debris forms clumps, which would be unlikely to form in dead tissue (Figures 13, 14 &17).

f)        The biopsy. No clear degenerative changes can be identified in the biopsy core specimens. There are occasional foci where epithelial cells may be showing very early changes (Figures 16 & 17) but, in the context of the variations inherent in cancers, the changes are not definite, and there is certainly no evidence of the degeneration that results in loss of gland lumina.



Figure 18: Core biopsy specimen, showing a focus which might demonstrate some early degenerative changes (red arrows) in the carcinoma glands. There might be two small inflammatory cells in the stroma right of centre (blue arrow). (Objective x40)

{C}{C}{C}{C}{C}{C}{C}{C}{C}{C}                           FIGURE 19                {C}{C}{C}

{C}{C}{C}{C}Figure 19:  Core biopsy, showing some more borderline, possibly very early, degenerative changes (red arrows). Present, is one of the few  collections of inflammatory cells, probably lymphocytes, but occasional plasma cells have been identified (blue arrows) in loose stroma, separated from the malignant gland epithelium. (Objective x40)




A                   Doubling Time. The prostate cancer studied here had a (PSA) doubling time of ~4.75 year. There should be some difficulty imagining cancer cells lying dormant (in G0/1 phase) for some four years and 275 days, before activating and progressing to S phase, G2 phase, and on to mitosis over the next day or so. A more plausible proposal would be that division and apoptosis proceed more swiftly, but that the balance very slightly favours proliferation, giving the net doubling time of 4.75 year. This is mentioned because the Lithium Rx effects may then be better understood. Just what such a cell doubling time would be is unknown.

B                   Prostate Size Changes. The make-up of the data-sources for assessing prostate size changes can hardly be regarded as optimum. The apparent decrease in global gland volume after 27 days would be difficult to explain on the basis of tumour reduction. A considerable amount of the gland is of fibromuscular tissue, with the cancer located in a limited percentage distribution. There are likely to be two forms of size reduction :

a)      A fluid shift, whereby extracellular fluid leaves the gland. Some support for this suggestion comes from more recent investigations into the renin-angiotensin system within the prostate[4]. The angiotensin II type 1 receptor (AT1-R) has been located on the peri-urethral stromal smooth muscles in one study and in glandular epithelial cells in another (there being some discrepancies). Since AT1-R belongs to the G-protein coupled receptor family, Lithium may influence angiotensin controls. (When the gland is examined microscopically, clefts between the smooth muscle bundles represent what must be a surprisingly large potential reservoir for extracellular fluid [see Figure 15]. Such a reserve might make sense given the likely rapid need for fluid when the gland is actively secreting for a short time, leaving a “space” outside the acini to accommodate their distension prior to discharge.) Prostate size changes, probably involving fluid shifts in the outer part of the general muscle bulk, may be more appreciable than generally considered.

b)      Tumour volume reduction. Assuming that tumour reduction occurs, it would be expected to be a later effect of Lithium Rx, localized to the tumour, and correlated with some cytological or focal structural changes. The oblique measurements generally support a reduction in the “right oblique” distance with respect to the width. The “left oblique” measurement in the gross surgical specimen seems spurious. One explanation for this may be that rigor mortis (as seen with erector pili smooth muscle) produced a muscle contraction at the left posterior, which could not occur on the right because of the tumour. The width may involve smooth muscle with less ability to shrink. Another explanation may be simple: the specimen was removed at ~22:00 h, had a paper towel wrapped around it and was placed in a formaldehyde-saline specimen pot. The gland may have been positioned so that it rested on the left posterior lobe overnight, with the paper holding it in position. On this basis, the left posterior aspect had fluid pressed from the extracellular spaces that are located between the smooth muscle bundles. If this were the case, pressure was gentle, because microscopy of the region still demonstrates many extracellular clefts (Figure 15). Other attempts at explanation may be difficult.




{C}{C}{C}{C}{C}{C}{C}{C}{C}{C}Figure 20: A section through the left posterior lobe with the surgical excision margin, marked with a blue dye, at the top. The larger walled spaces are of arterioles, venules and glands, the latter showing benign cystic changes. The smooth muscle bundles (red) are prominent and course between the spaces. The clefts and spaces between the smooth muscle are still prominent (green arrows). (Objective x4)

Examination of the histology preparations provides good evidence that a degenerative process (see later) was underway that was capable of causing a volume reduction of the tumour.

If there have been these volume changes as a result of the Lithium Rx, there may be a number of therapeutic ramification that can be considered and applied forthwith :

a)      Preoperative Lithium Rx may reduce the size of the prostate gland and may make surgical removal easier,

b)      If the Lithium Rx induced a later degeneration of the tumour, there may be good grounds to believe that the Rx may reduce the likelihood of tumour spread at or about the time of surgery, meaning that the cell metastatic potential may have been reduced even if there was no clear cytological change. Confirmation of this would require further work.


B                   Cytological Degeneration. Degeneration, with the hall-mark change of lumen loss, could not be detected clearly in the biopsy, despite 4 good cores (18-20 mm long) sampling tumour-affected gland regions on the right. Whilst this does not exclude the possibility that such degeneration was missed in the sampling, this possibility seems unlikely. Given the patchy nature of the degeneration, it probably started several weeks before the prostatectomy: attempts to identify early changes in the biopsy were inconclusive (Figures 18 & 19). There is no direct proof that the cellular degeneration was as a direct result of the Lithium Rx. The assumption here is that it probably was, in which case it may have taken some 10 weeks or so to develop in patient PC, who had the longer doubling time (see earlier). Patient LC had a much faster growing tumour, and there is a likelihood that a similar degeneration developed, probably in ~30 days.

In the case of LC, speculation was that the Rx damaged the tumour stem cells. Such an explanation may still be applicable for patient PC. A clearer idea of the effects of the Rx may await trials on prostate cancer patients with disseminated disease whose treatment options have run out.

If the Lithium Rx can produce degeneration in prostate cancer deposits, it may be applicable for the treatment of symptoms related to space occupation or pressure, although it may have a long lag time.

C                   Degeneration timing. All cancers show varying degrees of cellular degeneration or regression in their variegated patterns. What is important here is to establish that the cellular degeneration noted in the operation specimen was not simply the “background” level that existed from before the biopsy. There are good reasons to believe that the degeneration that has been noted in the operation specimen is not represented, both in type and extent, in the biopsy specimens, consistent with the conclusion that the described degeneration was recently acquired after the biopsy, but before the prostatectomy operation :

a)     The degeneration could not be identified clearly in the biopsy specimens (Figures 18 & 19). There is the possibility that the sampling missed areas with degeneration, but this event, whilst possible, would seem unlikely,

b)     Consistent with the negative biopsy findings, was the appearance in the surgical specimen of what appear to be transition steps from viable-appearing cancer cells to those with the degenerative features that have been described (Figures 13, 14, 16 & 17). That would place the dating of the onset days to a week or so before the operation.

c)      The degenerative changes are patchy (Figures 10 & 13), with most showing rather similar features, as though the areas involved are early examples of the degeneration starting at, or about, the same time

d)     Features suggesting old stages in degeneration are not often seen; likely examples are the clumping of masses of nuclear material in the blocked lymphatic channels (Figures 13, 14 & 17)

e)      Inflammatory cells are generally scanty or mild, with no clear relationship to the degenerative process. This is a feature of early apoptosis; later stages in apoptosis may set up inflammatory responses.

The conclusion from all of the above, is that the degeneration probably started after the biopsy, perhaps days to a week or two prior to the surgery. With that in mind, consideration must now be given to the possibility that there was a spontaneous regression quite unrelated to the Lithium Rx.


D                Spontaneous Regression. There can be little doubt that malignant tumours can spontaneously regress, either partially or completely. But it is generally very rare. Tumours with particular notoriety in this respect are those derived from the neural crest: malignant melanoma[5],[6] and neuroblastoma[7]. The possibility of a prostate cancer spontaneously regressing here, whilst possible (and reported[8]), would seem extremely unlikely. The mechanism of spontaneous regression is generally considered to be an immune process. In the case of human malignant melanoma, CD4 T lymphocytes CD8 T lymphocytes are conspicuous in the early stages of remission. However, a recent study of gene responses in an experimental swine regressing melanoma[9] casts some doubts, in that genetic indicators of early regression precede the cellular immune responses. The early genes up-regulated were concerned with immune responses and cell death; those down-regulated were for cell cycle and DNA replication, recombination and repair.  The earliest immune cell gene responses were for the monocytes (~day 49), then T/NK (~day 70). The humoral/immunoglobulin responses were early (~day 49). In particular, when immune cells did appear in the tumour, CD8 T cells exceeded CD4 T cells. Others place importance on CD8 T lymphocytes and NK cells in tumour rejection, with inhibition of these induced by Treg cells[10]. If Lithium Rx has induced a “spontaneous regression-like” change at the gene expression level, there may have been little early immune cell involvement, hence the paucity of such cells. Considerable interest would be justified in examining the processes involved, particularly at the gene level.

If the Lithium Rx has induced the degenerative changes, it would be unique amongst the anti-cancer drugs, because Lithium is not known to have any direct

toxic effect upon cancer cells at physiological levels.

E                   Prostate Specific Antigen levels. As with the tumour marker observed with patient LC, changes in the PSA levels (if real) were very modest at best, and only can be regarded as possible trends because of the inherent variations introduced by day-to-day patient and laboratory differences: the Rx may have caused a fall in the PSA level, possibly due to early degeneration of the progeny glandular cells. There may be a later rise, possibly because of breakdown of the progeny glandular cells and release of the PSA, as occurs in tumours after other forms of destructive treatments. The results can only be used for planning observations in the future. Based upon the hypothesis with patient LC, the markers were believed to reflect more the progeny cell numbers – they were smaller, not actively proliferating and occasionally breaking up, but still generally productive of the marker. This hypothesis could only be applied early in the cause of events for PC. He would have to have waited much longer in order to assess the ultimate effects upon the marker.

F                    Lymphocyte Surface Markers. These showed appreciable changes, but provide only circumstantial evidence for usefulness: the histology of the tumorous areas showed remarkably few inflammatory cells of the lymphocyte/monocyte type. Perhaps effector cells have not yet been activated, or perhaps they moved rapidly into the epithelial layer to participate in the degenerative change, so that they could not be identified by the staining method used here; but no evidence could be identified to support this. Likewise, there was minimal evidence of any reactive inflammatory process; a feature that could support the belief that the degeneration was an apoptosis-like process, or that the Lithium Rx blocked an inflammatory response. With Lithium Rx, both patients LC and PC showed appreciable stimulation of the NK cells numbers in their peripheral blood. Lithium Rx would seem to be one of the few agents that may be able to do this, although co-stimulant(s) from the tumour may be necessary as well. This means that the bone marrow and lymph nodes for NK proliferation are not suppressed[11]. Patient PC showed a healthy increase in CD8 T cells (which include CTL cells). After an initial fall, the CD4 T cells then rose, matching the CD8 T cells rise. These would indicate that the bone marrow and thymus production of both cell subtypes is not appreciably suppressed but, instead, stimulated. A single determination of Tregs showed no clear change one way or the other. If Tregs are affected, the change may only be manifest within the target tissue – the prostate cancer. Perhaps the surprisingly small number of lymphocytes found in the tumour may indicate an exclusion or destruction of Tregs in situ. Immunological decoration of the TILs, looking for Tregs and the other CTLs (if possible) might clarify this.

G                  Nuclear Volume. Attempts were made to assess the volume changes of the cancer cell nuclei. Images from objective x40 were printed to 19.5 x 28.5 cm. The larger, vesicular nuclei with no overlap were identified and measured in millimetres, collecting measurements from 20 nuclei (usually). The “volume” was then calculated by - - - length x (width)2 x 0.524, to express the “volume” in mm3, figures that have been presented in the histogram. The result is crude, with selection and no allowance for nuclear orientation and shape irregularities, but may give some rough indication of trends.




Figure 21: The alphabetical identities are:

A  Biopsy; “Normal”/PIN, Left lobe - lavender

E  Biopsy; Right upper lobe - light green

B  Biopsy; Cancer, Right upper lobe - brown

F  Surgical; Right upper lobe - red

C  Biopsy; Cancer, Right upper lobe - yellow

G Surgical; Right upper lobe - pink

D  Biopsy; Cancer, Right upper lobe - blue

H Surgical; Right upper lobe - dark green

The sample from the left lobe biopsy (A) showed small nuclei, as expected. Those from the right lobe biopsy (B, C & D) showed similar modes and spread, with an increase over the more normal sample. One sample from the biopsied right lobe (E) showed an unusually wide scatter, with a mode larger than those sampled from the surgical specimen. The results from the surgical specimen, right upper lobe (F, G & H) show similar modes, generally larger than those from the biopsy.


The nuclear “volumes” from the biopsy and the surgical specimen show interesting changes. However, they can only be used as indicators of trend. That is because the assessments are too crude for statistical analysis. The trend is towards larger nuclei during the time interval between the biopsy and surgical excision. One sample from the biopsy (E) may reflect early degenerative changes, but this conclusion is not provisional.



In this Chapter, two patients with cancer have been presented and contrasted. Both were treated with intermittent, bolus Lithium on alternate days as the only form of therapy. The patient with the lung cancer (LC) enjoyed an appreciable size reduction of her tumour, and has been presented in an earlier Chapter; the patient with prostate cancer (PC), whose study comprised the bulk of this Chapter also enjoyed effects which could be beneficial.

While there may have been tumour volume reductions for both patients, which could be of benefit if there were symptoms from space occupation or pressure, these cannot necessarily be translated into improved survival. If the tumour stem cells are relatively unaffected, the Lithium Rx could be merely a method of obtaining a medical debulking, without any real survival advantage. If the tumour stem cells are suppressed, there may be an increased survival. Further research is required.

Given that Lithium carbonate is very cheap and, in this context, relatively safe, provided that basic precautions are taken, such as checking renal function and serum calcium prior to treatment initiation, then measuring the serum Lithium, calcium, renal function and thyroid function from time to time, should provide adequate monitoring. The biggest problem may be patient compliance; persuading the patients that the Lithium is not to treat a psychiatric disorder and emphasizing with them to take the bolus doses strictly in accordance with the protocol: if a dose is missed, not to take a double dose the next time, and to be sure to miss all doses on the alternate, non-treatment days, and maintain a good fluid through-put. Both patients received 500 mg Lithium carbonate six hour apart on the treatment day, had a day without treatment, then a repeat of the bolus doses on the next treatment day, and so on. Patient PC could have had a higher dose regime. Provided the monitoring checks are performed, the Lithium Rx may be continued indefinitely.

A latent interval of ~30 day probably ensued before any appreciable response in patient LC, with a fast growing tumour. For patient PC, the latent period was greater than ~3 weeks, and was not fully expressed at day 84, the day of the prostatectomy. There are indications that effects probably started about day 60-70.

The volume of cancer cell nuclei seemed greater in the prostatectomy specimen. This does raise the possibility that, in patients with bone pain due to metastasis, there may be a “flare” response during the latent interval after starting Lithium Rx – being an increase of pain preceding an improvement.

Whilst there were interesting changes to the lymphocyte sub-populations, there was little clear evidence that lymphocytes or NK cells were involved appreciably in the tissues of the prostate. Accordingly, the mode of action is unclear, with the possibility that the Lithium Rx involvement may be at the level of gene expression[12].

On current findings, starting a Lithium Rx course before a planned prostatectomy may reduce the gland size and also reduce the risk of peri-operative metastases.



Malcolm Adams Traill

Copyright © MA Traill 2008-10-27  




[1] Chapter “Hyperthermia, Lithium and 434 MHz UHF;” in www.malsmusings.info

[2] Levine S Saltzman A. Immunopharmacology 1991; 22:207-213

[3] Traill MA. IRCS (Res. Clin. Pharm. Ther.; Metab. Nutrit.; Psych. Clin. Psych.) 1974; 2:1055

[4] Chow L Rezmann L et al. Mol Cell Endocrinol. 2008; doi:10.1016/j.mce.2008.08.032

[5] Emanuel PO Mannion M et al. Am J Dermatopath. 2008; 30(2):178-181

[6] Wenzel J Bekisch B et al. Am J Clin Path. 2005; 124:37-48

[7] Guin GH Gilbert EF et al. Am J Clin Path. 1969; 51(1):126-136

[8] Schurmans JR Blijenberg GH et al. J Urol. 1996; 155:653

[9] Rambow F Piton G et al. Neoplasia 2008; 10(7):714-726

[10] Ghiringhelli F Menard C et al. Cancer Immunol Immunother. 2007; 56:641-648

[11] Caligiuri MA. Blood 2008; 112:461-469

[12] Chapter “LITHIUM and NON-PSYCHIATRIC DISEASE ;” in www.malsmusings.info