Cancer Chronotherapy

The treatment of cancer is usually scheduled by the convenience of the health care system first, according to the availability of the clinical facilities and of the physician or nurse. Chemotherapy being usually highly toxic, the drug may cause harm to various organs, from the gut and kidney to the heart and/or bone marrow. The cancer (shown as a crab) may also be hit, but the treatment by convenience may not be optimally planned. As a consequence, before killing the cancer, the treatment may kill the patient (Figure 3/I, #1). Charts such as those shown in Figure 1/IV are helpful for determining when a given agent is less toxic to the host; they make it possible to time treatment to minimize the undesired toxic effects of the treatment. This targeting for the times of the optimal tolerances constitues an important improvement, but in itself it is not a sufficient advance. If the scheduling of the treatment takes into consideration only toxicity to the host, treatment may not be optimal in terms of killing the cancer. As a result, targeting by host markers lessens the impairment of life quality by debilitating and nauseating drugs, but it does not necessarily increase the treatment's efficacy (Figure 3/I, #2). The possibility of minimizing toxicity by timing is illustrated for the case of adriamycin (Figure 3/II). Results from five separate studies involving a total of 858 mice show that mortality reaches 80% when adriamycin is administered in the middle of the daily dark (active) span, whereas it is only 30% when given in the daily light (rest) span. Overall, the time when this drug is least tolerated is late during the dark (active) span. Translating these results to humans, some usual clinic hours may correspond to a time when the host is most susceptible to the toxicity of this drug. It seems reasonable to seek the time of best tolerance for treatment unless the best time for killing the cancer can be determined.

Large-amplitude circadian (and other) rhythms have been mapped for some tumor markers in saliva and urine, where they may not or at best only indirectly reflect tumor burden. The non-invasive assessability in serial samples of urinary or salivary rhythms, and any immediate decrease as an index of the time of drug activity could render these markers suitable for guiding treatment timing so as to optimize efficacy. Marker rhythm-guided chronotherapy was carried out with prednisolone on the LOU rat bearing a transplanted immunocytoma: the excretion of light chains by this tumor is circadian periodic for a large part of the lifespan remaining after tumor inoculation. In this very attractive model, a therapeutic gain of about 70% is associated with optimal timing in relation to the circadian rhythm in urinary light chains excretion used as a tumor marker (Figure 3/III, left).

In the clinic, a relatively unspecific marker, namely cancer temperature, served for guiding the radiotherapy of patients with large tumors of the oral cavity (Figure 3/III, middle, right and bottom left). Radiation was applied for five weeks. Patients were randomly assigned to receive daily treatment at one of five circadian stages, either at the time of their daily peak tumor temperature (shown by a star) or 4 or 8 hours before or after that time. Peak tumor temperature was determined by assessing the circadian variation from repeated measurements of the tumor temperature, taken several times a day for a few days prior to the start of treatment. Tumor regression rate was largest for those patients receiving the treatment at the time of their peak tumor temperature. Patients treated at that time also had the largest percentage of disease-free survival at a two-year follow-up. Both in the experimental laboratory and in the clinic, chronotherapy is feasible and accompanied by large therapeutic gains. While targeting treatment in time by tumor markers may increase the chances of killing the cancer, the treatment may still be accompanied by great toxicity (Figure 3/I, #3).

The longitudinal assessment of proliferation markers in a patient (EH, 72 y) with a müllerian duct adenocarcinoma involving the ovary reveals circadian, infradian (notably circaseptan) and ultradian components, notably with periods of about 14 to 16.8 hours, as illustrated for the salivary concentration of CA130 and for the urinary excretion rate of macrophage-colony stimulating factor (M-CSF) (Figure 3/IV). When, apart from their circadians and infradians, some tumor markers exhibit ultradian variations, and thus more than one peak per day, the best time can be sought to administer the treatment to optimize its efficacy first while also, as a secondary consideration, attempting to optimally shield the host from the treatment's toxicity (Figure 3/I, #4.)

Those not statistically inclined may proceed directly to chapter 5.