Cancer chronomics III. Chronomics for cancer, aging, melatonin and experimental therapeutics researchers

Abstract

This position paper documents the merit of including for basic and clinical investigations the mapping of circadian and other rhythms and yet broader chronomes, time structures in and around us. Chronobiometry used herein relies on inferential statistical methods and on materials documented earlier. The circadian amplitude of melatonin is shown to relate both to cancer risk and to the presence of overt cancer, when no differences are found in the 24-hour average of melatonin. Optimization of treatment by timing, thoroughly documented along the circadian scale earlier, could be broadened to include optimization along the scale of the week, and eventually beyond. In both cases, reliance on marker rhythmometry is advocated. More generally, the limits of knowledge are expanded by considering already mapped spectral components and their characteristics that can be influenced by the dynamics of heliogeomagnetic signals heretofore unassessed.

Figure 1

Figure 1a. Least-squares fit of 24-h cosine curve to urinary melatonin excretion documents circadian variation and assesses rhythm characteristics as point-and-interval estimates amenable to inferential statistical testing. Please note (middle) that there is no up- or down-regulation; there is no difference in MESOR (P = 0.431), in keeping with the Guernsey III study (4). By contrast, for the difference in amplitude, the relatively small sample notwithstanding, the likelihood that the difference is due to random sampling error is small: 0.022 (5). © Halberg. Figure 1b. In the absence of a difference in MESOR of circulating melatonin, in the face of no up- or down-regulation, a difference in circadian amplitude is found between a group of 39 cancer patients and that of 28 healthy subjects. Cancer patients have numerically lower circulating melatonin concentrations during the rest span but statistically significantly higher melatonin concentrations during the active span (around noon). Inter-group difference in MESOR is not statistically significant (NS). Note further a P of 0.032 for a difference around noon, which reverses sign around midnight (–9). © Halberg. Figure 1c. Cosinor display of circadian rhythm in circulating melatonin concentration in cancer patients by comparison with healthy subjects. © Halberg.

Figure 1

Figure 1a. Least-squares fit of 24-h cosine curve to urinary melatonin excretion documents circadian variation and assesses rhythm characteristics as point-and-interval estimates amenable to inferential statistical testing. Please note (middle) that there is no up- or down-regulation; there is no difference in MESOR (P = 0.431), in keeping with the Guernsey III study (4). By contrast, for the difference in amplitude, the relatively small sample notwithstanding, the likelihood that the difference is due to random sampling error is small: 0.022 (5). © Halberg. Figure 1b. In the absence of a difference in MESOR of circulating melatonin, in the face of no up- or down-regulation, a difference in circadian amplitude is found between a group of 39 cancer patients and that of 28 healthy subjects. Cancer patients have numerically lower circulating melatonin concentrations during the rest span but statistically significantly higher melatonin concentrations during the active span (around noon). Inter-group difference in MESOR is not statistically significant (NS). Note further a P of 0.032 for a difference around noon, which reverses sign around midnight (–9). © Halberg. Figure 1c. Cosinor display of circadian rhythm in circulating melatonin concentration in cancer patients by comparison with healthy subjects. © Halberg.

Figure 1

Figure 1a. Least-squares fit of 24-h cosine curve to urinary melatonin excretion documents circadian variation and assesses rhythm characteristics as point-and-interval estimates amenable to inferential statistical testing. Please note (middle) that there is no up- or down-regulation; there is no difference in MESOR (P = 0.431), in keeping with the Guernsey III study (4). By contrast, for the difference in amplitude, the relatively small sample notwithstanding, the likelihood that the difference is due to random sampling error is small: 0.022 (5). © Halberg. Figure 1b. In the absence of a difference in MESOR of circulating melatonin, in the face of no up- or down-regulation, a difference in circadian amplitude is found between a group of 39 cancer patients and that of 28 healthy subjects. Cancer patients have numerically lower circulating melatonin concentrations during the rest span but statistically significantly higher melatonin concentrations during the active span (around noon). Inter-group difference in MESOR is not statistically significant (NS). Note further a P of 0.032 for a difference around noon, which reverses sign around midnight (–9). © Halberg. Figure 1c. Cosinor display of circadian rhythm in circulating melatonin concentration in cancer patients by comparison with healthy subjects. © Halberg.

Figure 2

In data taken from a published graph (11), a smaller circadian amplitude can be demonstrated in older subjects when the data have been log₁₀-transformed (right), but not when they are expressed in original units (left) (10). In this case, the transformation reveals to the naked eye the inter-group difference by daytime. © Halberg.

Figure 3

The same total dose of the same molecule, lentinan, depending on the circadian and circaseptan patterns of its administration, stimulates the growth of a malignancy on a conventional administration pattern while it inhibits the malignancy with sinusoidal administration (1). An undesirable harmful effect becomes a desirable one by accounting for chronomics. © Halberg.

Figure 4

Blunders (bottom) due to the fact that two rhythms are out of phase. Top: two sinusoids, A and B, with the same period and the same average representing two rhythms being compared. Curves are in antiphase, one peaking when the other has a trough. At the midline crossing, there will be no difference, as shown by two columns below. At the peak of A at 06:00, A>B, and the opposite will be found by somebody checking at 18:00. By integrating over 24 h, such blunders can be avoided, but information on the behavior of rhythm characteristics such as amplitudes (Figure 1 and Figure 2) is lost (22). © Halberg.

Figure 5

Artifacts (bottom) arising when sampling on two rhythms with a different period at arbitrary times (top). Example shows a big difference in period but various similar, including opposite artifactual differences will come about more slowly with any smaller difference in period, including an about half-hour difference in the period of the body core temperature rhythm after blinding in the earliest 1950s that led to the recognition of free-running (22). © Halberg.

Figure 6

Time-dependence of β-ATP peak in L1210 leukemia cells (after detrending original data of W. Ulmer) (1). Note that circaseptan-to-circadian amplitude ratio is larger than unity. Three other cell cultures also exhibited a prominent circaseptan component and statistically significant circadian and about-monthly rhythms of smaller amplitude than the circaseptan (not shown). The circaseptan-to-circadian amplitude ratios are readily seen to vary between 2.5 and 4, Table 1. Nonlinear analyses yield period estimates with 95% confidence intervals covering 24 hours and 7 days, respectively, so that a free-run cannot be validated. Chronobiologic serial sections exhibit a stable phase throughout the whole observation span. Results pursued in further research by W. Ulmer with G. Cornélissen (33, 34). © Halberg.

Figure 7

Spectrum of the detrended data shown in Figure 6 illustrates the relative prominence of the about-weekly versus the about-daily and about-monthly components. Results pursued in further research by W. Ulmer with G. Cornélissen (33, 34). © Halberg.

Figure 8

Time courses of the frequency structures of the speed of the solar wind (SWS) (top) and of an elderly man’s (FH) systolic and diastolic blood pressure and heart rate, SBP, DBP and HR (rows 2–4, respectively), examined by gliding spectral windows. Human SBP selectively resonates with SWS (top 2 sections). Much less, if any resonance, only minor coincident change with DBP or HR (bottom 2 sections). Aeolian rhythms in gliding spectra of SWS and SBP change in frequency (smoothly [A] or abruptly [B,C,D], bifurcating [D,F] and rejoining [G], they also change in amplitude (B) (up to disappearing [C,E] and reappearing). During a nearly 16-year span there are no consistent components with a period averaging precisely 1 year in the 3 physiologic variables, probably an effect of advancing age. While post hoc ergo propter hoc reasoning can never be ruled out, an abrupt change on top in SWS is followed in the second row in SBP by the disappearance of some components, suggesting that as a first demonstration, some of FH’s cis- and transyear components were driven by the SW [their disappearance with a lag of about a transyear following the disappearance (subtraction) of the same components from the SWS spectrum may indicate resonance for an additional cycle]. The persistence of other spectral features in turn suggests endogenicity, i.e., an evolutionary acquisition of solar transyear oscillations that may reflect solar dynamics for the past billions of years. Blood pressure and heart rate data are from a man 70 years of age at start of around-the-clock monitoring, mostly at 30-min intervals for nearly 16 years, with interruptions (N=2418 daily averages, total about 55000). Gliding spectra computed with interval = 8 y, resolution low in time but high in frequency, increment = 1 month, trial periods from 2.5 to 0.4 y, with harmonic increment = 0.05. Darker shading corresponds to larger amplitude. When several of these broad bands disappear in the SWS, at E, parts of the bands in SBP also disappear, with a lag (delay) at E’, while other parts of the originally broader band persist. These components are presumably built into organisms over billions of years, as persistence without corresponding components in SWS shows, but can be driven in part by the solar wind, as their disappearence after loss of corresponding components in SWS suggests. “Aeolian”, derived from Aeolus, Greek God of winds, who packed the winds up and then let them loose and had them change. © Halberg.

Figure 9

Mitotic activity in the bone marrow of cancer patients is characterized not only by a circadian rhythm but also by a circannual variation. The waveform of both components is shown by stacking the data over an idealized cycle (top). Cosinor displays (bottom) readily convey the prominence of these components and the timing of their overall high values recurring in a cycle. Data of M Blank (40). © Halberg.