From www.bloodjournal.org by guest on June 16, 2017. For personal use only. DNA Synthesis in Human Bone Marrow Is Circadian Stage Dependent By Rune Smaaland, Ole D. Laerum, Knut Lote, Olav Sletvold, Robert 6.Sothern, and Robert Bjesknes Fraction of human bone marrow (BM) cells in DNA synthesis has been studied by sampling B M from the sternum or the iliac crests every 4 hours during one 24-hour period in 16 healthy male volunteers. Three of the subjects underwent the sampling procedure twice, resulting in 19 24-hour profiles. The percentage of cells in DNA synthesis measured by flow cytometry demonstrateda large variation along the circadian time scale for each 24-hour profile, with a range of variation from 29% to 339% from lowest to highest value. Seventeen profiles (89.5%) had the highest DNA synthesis during waking hours between 08:OO hours and 2O:OO hours, and the lowest percentage of cells in DNA synthesis between 0O:OO hours and 0 4 : O O hours. TKe mean value of the lowest DNA synthesis for each 19 2ehour period was 8.7% & 0.6%. while the mean value of the highest DNA synthesis was 17.6% 2 0.6%, ie, a twofold difference. There was no difference in DNA synthesis between winter and summer. A significantly higher DNA synthesis was demonstrated for samples obtained from sternum as compared with the iliac crests, but the same circadian pattern was demonstrated for both localizations.By taking circadian stage-dependent variations in DNA synthesis into account it may be possible to reduce B M sensitivity to cytotoxic chemotherapy, to increase the effect of hematopoietic growth factors as well as increase the fraction of proliferating cells with careful selection of time of day for harvesting B M cells for auto- or allografting. o 1991 by The American Society of Hematology. B To our knowledge only two studies measuring the DNA synthesis in human BM according to circadian stage have so far been r e p ~ r t e d , ' ~in. ~one and four individuals, respectively. Therefore, there has been an urgent need for a more extensive study of a possible temporal variation in proliferative activity of the human BM. If large enough, such temporal variations in BM cell proliferation could be of clinical importance both relative to optimization of cytotoxic therapy and administration of hematopoietic growth factors. The selection of time of day for harvesting BM cells for auto- or allografting could possibly also be optimized. We have conducted a study investigating the DNA synthesis in human BM cells sampled several times during 19 24-hour periods in 16 healthy male subjects. BM cells were aspirated by a standard technique used in the clinic, and analysis of DNA content has been performed by flow cytometry. ONE MARROW (BM) suppression is commonly associated with cytotoxic treatment of cancer, and is generally seen following combination therapy using different cytotoxic drugs.'.' It represents a major problem in cancer chemotherapy, because therapeutic response usually requires drug doses inducing BM hypoplasia. The cytotoxic effect on the BM is due to a potentially irreversible damage of pluripotent stem cells, early committed progenitor cells, and proliferating cells later in the maturation process, as well as to regulatoly stroma cells in the BM microenvironment? This sensitivity to cytotoxic therapy is to a great extent related to the high proliferation rate of BM cell~,4.~ although other mechanisms may be involved as well. Acute BM suppression may not only lead to serious infections, but also to dose reductions and postponement of treatment courses, as well as reduced duration of useful treatment. In addition, the possibilities of treatment in the event of relapse may be reduced. It is well documented that the susceptibility to cancer chemotherapy shows circadian variations in laboratory animals."8 In addition to reduced mortality due to acute toxicity, it has also been shown that an increase in tumor effect or cure rate can be ~ b t a i n e d , ~ or . ~ .that ' ~ it is possible to eliminate or reduce drug-induced death due to toxicity, while still using an effective dose.I4 Circadian and circannual variations in proliferative activity in murine BM, both regarding colony-forming unit granulocyte-macrophage (CFU-GM), CFU-spleen (CFU-S), and DNA synthesis, have also been shown."-*' In addition, clinical studies have demonstrated a circadian dependence of cytotoxic drugs to BM toxicity, showing less dose reductions, less treatment related complications, and less postponements of drug courses when drugs have been administered at certain There are also a few clinical studies either demonstrating or suggesting a reduced chance of relapse as well as increased long-term survival when cytotoxic therapy has been administered at specific times of the These time-dependent variations in toxicity and survival have not been generally recognized in practical-clinical treat~nent.'~ This lack of recognition may partly be due to the fact that there are few data on directly measured biologic rhythms of proliferative parameters in human BM. Blood, Vol77, No 12 (June 15), 1991: pp 2603-2611 MATERIALS AND METHODS Subjects. From November 1986 to August 1988 we obtained BM samples from 16 healthy male volunteers (mean age = 33.7 years; range 19 to 47 years) during 21 24-hour periods, ie, five subjects underwent the sampling procedure twice. To find out if the study was feasible, practically and ethically, the investigators started out sampling on themselves. Therefore, two of the first volunteers were MDs who had to do night-work. These two From The Gade Institute, Department of Pathology, Department of Oncology, and Department of Pediatrics, Haukeland Hospital, University of Bergen, Bergen, Norway; The Geriatric Department, The Deaconess Hospital, University of Bergen, Bergen, Norway; and The Rhythmometry Laboratory, University of Minnesota, Minneapolis. Submitted December 13, 1990; accepted February 13, 1991. Supported by the Norwegian Cancer Society and Michael Irgens Flocks Legacy. RS. is a fellow of the Norwegian Cancer Society. Address reprint requests to Rune Smaaland, MD, The Gade Institute, Dept. of Pathology, Haukeland Hospital, University of Bergen, 5021 BeKen, Norway. The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. section I734 solely to indicate thisfact. 0 1991 by The American Society of Hematology. 0006-4971J91/7712-0015$3. OOJO 2603 From www.bloodjournal.org by guest on June 16, 2017. For personal use only. 2604 subjects were omitted in the final analysis. After a pilot study of five individuals, it was found acceptable to include more subjects in the study, which was approved and performed in accordance with the guidelines of the regional medical ethics committee. All volunteers gave their informed written consent to enter the study, and all individuals included in the analysis followed their regular diurnal activity schedule with sleep at night for at least 3 weeks before the experiment. The subjects continued their usual activities during the study period in between times of sampling. They went to sleep after the 0O:OO hour sample was taken, and were awakened once for the 04:OO sample. Their diurnal rhythm was validated by determination of the cortisol level at every sampling point, which showed the usual circadian pattern for all individuals, ie, high morning levels and low evening levels. Protocol. Following periost anesthesia, BM was obtained by puncturing the sternum or one of the anterior iliac crests every 4 hours during a 24-hour period. To reduce the possibility that the repeated puncture procedure itself would interfere with the results, the start of the experiment was randomized to either 08:00, 1200, or 16:00hours, with the first time of sampling repeated at the end of each study for a total of seven samplesiprofile. The sequence of sampling from the three different anatomical sites was also randomized. No premedication was given. To exclude that any variations found could be attributed to sample dilution caused by local bleeding at the puncture site, differential counts were performed on smears from all individual samples. No samples had to be discarded because of unacceptable large peripheral blood admixture, ie, all smears were characteristic of BM (results not shown). Venous blood was also obtained from the subjects at the same time as BM sampling to determine hematologic parameters (total and white differential blood cell counts) and cortisol measurements. The blood was obtained as the initial procedure or immediately after the anesthesia of periost before the BM puncture. In this way an artificially increased level of cortisol resulting from the puncture procedure itself was avoided.” Procedure for BM sampling and sample handling. The puncture site was infiltrated with a local anesthetic (Lidocain, 20 mgiml; Astra, Sweden). No other premedication was administered. After BM (0.2 mL) was aspirated into a 2-cm3 syringe, one part of the sample was used for routine smears, while one droplet was stained directly for DNA flow cytometry (direct staining). Another droplet was placed onto each of two tilted microscope slides, to let the blood run down, thereby possibly increasing the fraction of marrow elements, which were immediately removed from the cover slides by a thin blade (made wet beforehand) of a knife and stained (indirect staining). Thus, two parallel samples from the same site were stained at each timepoint. Both samples of BM cells were added to 2 mL of ice-cold staining solution consisting of ethidium bromide, detergent, and RNAse according to the method described by Vindel@vv.’*The tubes were sealed and the solution shaken before being placed in an ice bath for at least 10 minutes. Flow cytometly. Both single cell suspensions were analyzed on a Cytofluorograph 50 H (Ortho Diagnostic Systems, Inc, Westwood, MA), interfaced to a Model 2150 Computer (Ortho). In the cytogram obtained, both the peak and the area of the red fluorescence signal were used for region-setting to discriminate the (G1 + GO) doublets from the real G2 + M cells. Thus, the second peak of the DNA histogram contained only the G2 + M cell population. This procedure was performed because the G1 + GO doublets may “contaminate” the G2 + M peak in the DNA histogram, leading to errors in the relative distribution of the different cell cycle phases. The total number of cells analyzed for each sample was 3 to 4 x lo4. Computerized analyses of the cell cycle distribution in the histograms were performed using the constant function of the cell cycle analysis program, by which the SMAALAND ET AL + percentages of cells in the G1 + GO, S, and G2 M phases were ~alculated.”.~‘ The mean coefficient of variation (CV) of the DNA histograms was 3.3%. Evaluation of fraction of cells in DNA synthesis (S-phase) was performed by taking the mean value of the S-phase of the two differently stained samples at each timepoint. In addition, the direct, the indirect, and the maximum values at each timepoint were recorded to more thoroughly evaluate the variation along the 24-hour scale. The maximum value obtained at each timepoint was included in the analyses because it may possibly represent the BM sample with highest fraction of proliferative cells. Statistical analysis. Data were analyzed by Student’s t-test (two-tailed; paired t-tests used for paired analyses of groups) and one-way analysis of variance (ANOVA), using data both in original units and as percentages of the individual mean DNA synthesis. In addition, the individual data obtained for each way of evaluating the DNA synthesis phase were analyzed for circadian rhythm by a computerized inferential statistical method involving the fitting of a 24-hour cosine by the method of least squares (Cosinor analysi~)?~ The rhythm characteristics estimated by this method include the mesor (rhythm-adjusted mean), the amplitude (half the difference between minimum and maximum of fitted cosine function), and the acrophase (time of peak value in fitted cosine function). A Pvalue for rejection of the zero-circadian amplitude assumption was determined on each data series. While the cosinor method may not accurately represent the true characteristics of the actual timedependent variations if assymetries exist in a time-series,” the procedure is nevertheless useful for assessing the presence of peri~dicities.~’ Individual rhythm characteristics were summarized for the group by population mean c ~ s i n o r .Spearman ~~ rank correlation test was performed for testing the correlation between the direct and indirect method of analyzing the DNA synthesis. RESULTS Circadian and circannual variation of DNA Jynthesis. The value of fraction of cells in DNA synthesis of BM cells harvested at each timepoint showed a large variation along the circadian scale for all 19 24-hour periods (Table 1). This finding was not explained by a corresponding variation in distribution of proliferative cells as judged by differential count of the BM smears at each timepoint, because there was no direct covariation between these two parameters. The range of change from lowest to highest DNA synthesis value during the 24-hour spans varied between 29% and 339%, with a mean and median difference of 118.2% 18.4% and 102.9%, respectively. As shown in Table 1, 17 of 19 series showed the highest DNA synthesis between 08:OO hours to 20:OO hours according to the cosinor analysis, ie, during daytime hours or early evening, and correspondingly, the trough of DNA synthesis during late evening and night. A complementary analysis showed that 17 of the 19 sampling periods had a lower mean fraction of cells in DNA synthesis from 0O:OO to 04:OO hours as compared with the mean DNA synthesis from 08:OO to 20:OO hours (P < .OOOl), and the DNA synthesis in the time span from 20:OO to 04:OO hours was lower in 16 of 19 periods as compared with the DNA synthesis from 08:OO to 16:OO hours (P < .005). The mean value of the lowest and highest S-phase was 8.7% 2 0.5% and 17.6% 2 0.6%, respectively, ie, a difference of 102.3% or a twofold variation in DNA synthesis depending on the time of measurement. Illustrative DNA histograms for two different subjects at two * From www.bloodjournal.org by guest on June 16, 2017. For personal use only. 2605 BONE MARROW DNA SYNTHESIS RHYTHM Table 1. Circadian Variation in Fraction (%) of Cells in S-Phase and Result of Single Cosinor Analysis Data Limits (S-phase) Series Subject ID 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 AA FLJ MJ RK EK SF RB RBJ #2 IK BCS RS #1 RS #2 OH os #2 GW KL #1 KL #2 ODL #1 ODL #2 Age (v) Parameters of 24-h Cosine Fit: N of Data Low High ROC (%) Mesor 2 SE 7 7 7 7 7 6 6 7 6 6 7 7 7 7 7 7 7 5 7 9.3 8.4 5.2 10.6 7.7 6.8 10.4 9.2 9.2 8.1 8.4 6.7 9.3 14.4 5.0 6.9 9.3 7.0 14.0 17.9 19.2 22.8 18.4 18.4 14.4 14.4 22.6 13.7 13.5 14.9 20.5 19.6 18.5 18.7 14.0 17.2 16.2 19.7 91.9 129.9 338.5 73.9 139.0 111.8 39.1 146.4 49.2 66.7 77.4 206.0 111.4 28.9 276.8 102.9 84.4 131.4 40.7 14.1 f 1.1 11.9 t 0.8 13.8 t 2.1 14.4 f 1.3 12.5 t 0.7 11.2 f 1.5 12.9 f 0.5 14.1 f 1.7 12.4 f 0.6 10.9 f 0.9 12.5 t 0.7 12.9 t 1.5 13.7 t 1.8 16.4 t 0.4 12.4 f 2.0 11.9 f 1.0 11.9 f 1.1 12.7 f 2.1 16.2 2 0.9 19 23 24 25 28 30 31 31 31 33 34 35 35 39 39 42 42 46 47 Amplitude ? SE 3.2 2 1.4 4.1 2 1.1 4.1 2 2.8 0.9 2 1.8 4.2 5 1.0 2.5 5 2.1 1.6 5 0.7 5.5 5 2.6 1.6 5 1.0 2.0 2 1.2 2.2 t 1.0 4.3 f 2.2 0.3 f 2.6 2.3 t 0.5 4.2 2 3.0 1.1 5 1.6 2.5 f 1.4 2.7 2 3.3 0.4 2 1.2 Acrophase' 04:55 08:14 11:21 13:58 13:53 09:06 04:41 14:21 15:43 14:54 13:39 12:23 07:45 11:39 18:30 10:56 17:43 17:24 09:46 Abbreviation: ROC, range of change from lowest to highest value. *In hours and minutes after local midnight. different timepoints (daytime and midnight) are shown in Fig 1. Although almost all subjects had their highest DNA synthesis during daytime, differences in phasing along the 24-hour period between the subjects were observed, ie, the I 1 SubjectRS Subject BS 5 e a e 6 12.00 hours a 4 c Relative DNA content V 00.00hours - 4 c Relative DNA content Fig 1. DNA histograms for t w o different subjects for two timepoints along the 24-hour time scale (day and midnight). The two peaks (2C and 4C) in each histogram designate the GO G1-phase and G2 M-phase. The part of the histogram in between is the S-phase. The height (ie, the area) of the S-phase expresses the percentage of cells in DNA synthesis. + + time of highest and lowest DNA synthesis differed to some extent between the individual subjects. Six examples of individual circadian stage-dependent variations of fraction of cells in DNA synthesis are shown in Fig 2 to demonstrate the slightly different phasing and the magnitude of variation in intraindividual DNA synthesis. The individual mean S-phase value of the 24-hour sampling period varied from 10.9% and 16.6%, ie, a difference of 52.3%. Due to this interindividual difference, the data were also normalized and expressed as percentage of the mean value. When pooling the data for all subjects both relative to the mean and highest S-phase values, a consistent pattern was seen, with a statistically significant lower DNA synthesis around midnight as compared with the day (Fig 3). The rhythm characteristics for the different ways of calculating the DNA synthesis data are depicted in Table 2. Due to different phasing among the subjects, the difference between the lowest and highest values is smaller as compared with the individual values. As can be seen from Table 2, the circadian stage-dependent variation is statistically significant for all methods of evaluating the data, analyzed both by ANOVA and the Cosinor method. Because the time of sampling started either at 08:00, 12:00, or 16:OO hours, DNA synthesis for the pooled data over 32 hours was evaluated (Fig 4). This makes it possible to observe the DNA synthesis for two consecutive dayperiods, demonstrating highest values during daytime (with a reproducible dip at midday) and lower late eveninghight values in between. No difference in DNA synthesis between winter, ie, October to March (13.2 rfr 0.4; n = 13), and summer, ie, April to September (13.3 f 0.6; n = 8) was observed, including all 21 24-hour profiles. DNA Jynthesisaccording to staining method. The procedure of letting the blood component of the BM run down From www.bloodjournal.org by guest on June 16, 2017. For personal use only. 2606 SMAALAND ET AL 19 8 - 221 Subject FLI 18- f s 9 17m 'g 3 16- . d 15- 2 n 08 12 16 20 00 04 08 08 Time of t h e day (MET.) 12 16 20 00 04 08 Time of the day (M.E.T.) Subject RS 0J OJ 08 12 16 20 00 04 08 1 12 Time of the day (M.E.T.) l6 1 16 20 00 04 08 Time of the day (M.E.T.) 12 201 Subject EK Subject KL #I 2 I , 16 20 00 04 08 12 16 16 Time of the day (M.E.T) the cover slide and then analyzing the cell components remaining on the c0ve.r slide (indirect staining method) represents a method intended to increase the fraction of proliferating cells of the BM aspiration sample. This is a simple method to use both for conventional investigation of . , . 20 Fig 2. DNA synthesis variation along the 24-hour time span in six different subjects, sampling of BM being performed every 4 hours (N = 7 sampleslsubject). Results are expressed as the mean of two parallel analyses. The time of starting the experiments was randomized to 08:OO. 12:OO. and 16:OO hours (M.E.T. = mean European time). Id * . 00 04 OS 12 Time of the day (h1.E.T) 16 BM smears and for more specific investigations of a purer BM sample. Nearly the same pattern of circadian variation was seen for the two ways of staining the cells, ie, the direct and the indirect method (Fig 5). The fraction of cells in DNA From www.bloodjournal.org by guest on June 16, 2017. For personal use only. 2607 BONE MARROW DNA SYNTHESIS RHYTHM Mean Values (Percent of Mean) Mean Values (Original Units) 15.5 f 2 14.5 -1 T 1201 T P 13.5 n Er 12.5 90 9 11.5 - ANOVA for Time Effect: F - 4 . 4 0 , ~4.001 0 10.5 12 16 Time of the day (M.E.T.) 08 04 00 20 I 00 20 Highest Values (Percent of Mean) Highest Values (Original Units) T Fig 3. Circadian variation in human BM DNA synthesis in 19 =-hour periods from 16 clinically healthy men (total N = 127). Timepoint means (absolute values and percentage) and standard errors of DNA synthesis are depicted along the 24-hour time scale. In addition, the highest DNA synthesis value measured of the two parallel samples is depicted correspondingly. 04 08 12 16 Time of the day (M.E.T.) T 1204 90 - -1 ANOVA for Time Effect: F=5.2S,ppcU.001 F- 4.68, p ~ 0 . 0 0 1 11 7 04 08 12 16 Time of the day (M.E.T.) 00 00 20 08 12 16 Time of the day (M.E.T.) 04 20 using the mean value of the two methods was 13.2%f 0.3% for the whole material. DNA synthesis according to anatomical localization. We found no statistical difference of the S-phase between the right (n = 46) and left (n = 44) iliac crests, with overall means being 12.2% f 0.5% and 13.1%& O S % , respectively (P= .16). A significantly highex S-phase was observed for samples obtained from the sternum (n = 51) as compared with the iliac crests (n = 90); 14.5%f 0.5% versus 12.6% f 0.3%,respectively (P = .0015). Comparison of timepoints by t-test showed a statistically significant difference between the two localizations for the samples obtained at 08:OO hours and 00.00 hours (P < .01 and P < .05, respec- synthesis was slightly higher for each timepoint when the indirect staining method was used. The difference was significant only for two timepoints, at 08:OO hours and 0O:OO hours; P < .01 and P < .001, respectively. However, when comparing the paired data available for all timepoints (n = 120), a highly significant difference was observed between the two methods, with a larger fraction of cells in DNA synthesis using the indirect staining method as compared with the direct staining method, 14.2% f 0.3% versus 12.7% 5 0.3%, respectively (P < .0001). A highly significant correlation was found between the two methods when comparing the two ways of BM sampling (r = .62; P < .OOOl). The fraction of cells in DNA synthesis when Table 2. Statistical Evaluation of Circadian Stage-DependentVariation of DNA Synthesis in Human BM Analysis by: ANOVA Variable Units N Mean value Mean value Highestvalue Highestvalue Direct method Direct method Indirect method Indirect method Original % o f mean Original % of mean Original % of mean Original % of mean 127 127 127 127 122 122 113 113 Arithmetic Mean t 13.20 k 0.32 100.0 k 2.2 14.342 0.36 100.0 2 2.2 12.50 0.34 100.0 2.4 14.16 2 0.37 100.0 2 2.3 * * SE Cosinor Population Mean Cosinor Summary: F P P Mesor ? SE Amp (95% limits) 0 3.70 4.40 4.68 5.25 2.70 3.69 2.29 2.45 ,004 ,001 <.001 <.001 ,024 .004 .051 ,038 ,006 ,004 13.06 2 0.35 ,003 14.14 1.54 (0.44, 2.67) 11.9 (3.7, 20.4) 1.79 (0.64, 2.96) 12.6 (4.9, 20.5) 1.36 (0.21,2.53) 11.6 (1.6, 21.7) 1.49 (0.03, 3.02) 10.0 (0.3, 20.4) 1300 1304 1312 1312 1212 1212 1340 1344 ,002 ,019 .022 ,045 ,042 0.36 12.49 & 0.44 14.05 k 0.43 (95% limits) (09:24, (09:32, (1O:OO. (10:04, (07:48, (07:52, (07:28, (07:40, 16:OO) 16:04) 16:04) 16:08) 15:56) 16:OO) 17:56) 17:52) Comparison of circadian results for different methods of estimation of DNA synthesis in 19 series obtained from 16 men. ANOVA, analysis of variance across all timepoints using all data in original units or as percentages of mean; Cosinor, summary of indiv+dual 24-hour rhythm characteristics by population mean cosinor using amplitude (Amp) in both oroginal units and percentages as of Mesor (Mesor, 24-hour rhythm-adjusted mean); Acrophase (0) reference, OO:OO, with sleepkest between 00:00-08:OO.The 95% limits for Amp and B giwm i f P I .@5from cosinor analysis. From www.bloodjournal.org by guest on June 16, 2017. For personal use only. SMAALAND ET AL .Tamplini Day 2 08 12 16 20 00 04 08 12 Time of the day(M.E.T.) 16 Flg 4. C i d k n varktion in human BM DNA svnthosb for 19 =hour p.riod.(total N = 127). Timepoint mwmand standard OWON am given. The time r u l e b extended for 32 hours due to the different times of starting each indMdualstudy 1text). tivcly) with marginal statistical significance at 1 6 0 (P= .07). To scc if thc higher proliferative activity in the samples obtaincd from thc stcrnum could in part explain the observed circadian rhythm, and to see if the rhythm at the two localizations was thc samc, we pooled all thc samplcs taken from stcrnum and iliac crests at cach timc point and analyzed cach sitc scparatcly. Duc to the intcrindividual diffcrcnccs in the mean S-phase, the data wcrc also normalizcd. A statistically significant effcct of time was found for the samplcs obtaincd from stcrnum whcn analyzcd as original valucs and as pcrccntage of the mcan, P = . 0 3 and P = .02 (ANOVA), rcspcctivcly. When analyzed .Sunphi Day 2 r 17 - ; ; IS - ; 16- I 14- ; I 13- ; I 12 - ; II - I I ; I IO- ; .knqIiinRDUy2 I OR 12 16 20 00 0.1 OR I ' 12 16 Time of the day(M.E.T.) Flg 5. Comprrhon oftmtiOn ofdh in DNA- Kcordlng to method of a ~ m n the g BM m p k . (direct W-1or i n d l m t [-&I), whh tho time rule extended for 32 hours. by cosinor we found P = .02 both for original values and as perccntagc of thc mean, with acrophasc at 1216 hours and 12:48 hours, respectively. The samc circadian pattcm was scen for the DNA synthesis in the samples from thc iliac crcst as for thc stcmum, although thc rhythm was lcss pronounccd. The P valuc was .Mwhen analyzing thc crcst data as percentage of the mean. Thc acrophasc was found at 1247 hours (Fig 6). The possibility that multiple punctures into the bonc could havc an impact on DNA synthesis by virtue of stress was considered. We therefore compared the level of the stress rclatcd hormone cortisol at thc start and at the end of thc sampling proccdure for cach individual, ie, 24 hours apart. No statistically significant diffcrcncc was observed (P = .97). Ncithcr was thcrc a statisticallysignificant diffcrcnce in perccntagc of cclls in DNA synthcsis samplcd 24 hours apart (P = .22). In addition, any possible stress rclatcd effcct on overall pattern of the poolcd data was minimized in the study protocol by starting sampling at threc diffcrcnt clock hours. DISCUSSION There is now increasing evidence that any compromise of dosage or dclays in treatment schedule diminish thc likelihood of canccr control or cure?'.ww On thc other hand, it sccms likely that treatment of some cancers would improve if doses of cytotoxic drugs could bc incrcascd.'" This is due to thc fact that the efficacy of most antineoplastic drugs is dosc intensity-dcpendcnt, ie, higher doscs ovcr a shorter timc span increase thc responsc rates and proportion of curcs." Thercforc, it hccomcs important to reducc the toxic effccts to normal sensitive tissues, especially the BM. Scvcral ways of circumventing this problem arc bcing cxplorcd, such as BM transplantation (auto- or allografting),"' use of recombinant human hematopoietic growth factors (granulocytc-macrophagccolony-stimulating factor (GM-CSF]/granulocytc-CSF [ G-CSF]):"." and rctroviral transfer of thc MDRl gcnc to primary hcmatopoictic progcnitor cclls (murinc studies)." In this study we suggest another approach, as we havc demonstrated large circadian variations in fraction of BM cells in DNA synthesis in each of 19 24-hour periods, which can be taken advantagc of whcn trying to optimize trcatment of canccr paticnts. A total of sevcn 4-hourly samples were obtained from each individual, with a median differcnce bctween the lowest and highest S-phasc valuc of 103%. Although there wcrc intcrindividual diffcrenccs in circadian phasing, a well-known phcnomenon for most rhythmic physiologic parameters, a consistent and statisticallysignificant pattcrn was manifested for the group as a whole whcn pooling the data from all subjccts. The pcriod of lowest DNA synthesis was found around midnight, and the highest DNA synthcsis was found during the day. This pattcrn was statistically validatcd for cach of four diffcrcnt ways of dctermining thc DNA synthesis at each timcpoint, as prcsentcd in Tablc 2. Bccausc the sampling period cxtended for 32 hours for the poolcd data, it was possiblc to graph the DNA synthesis over two consecutive daytimc periods. Higher DNA synthesis during daytimc on thc sccond day aftcr having bccn low From www.bloodjournal.org by guest on June 16, 2017. For personal use only. 2609 BONE MARROW DNA SYNTHESIS RHYTHM T T T T Fig 6. Circadian variation in DNAsynthesis according to ssmpling site {sternum [-.I or iliac crests [-O-]) along the 24-hour time scale. 01 Iliac Crests I - - 00 04 08 12 16 Time of the day(M.E.T.) during night corroborates the finding of a higher DNA synthesis during the daytime of the first day. However, due to different phasing between the subjects, the difference between the lowest and highest value was smaller than that found for the individual subjects. The average time of lowest fraction of cells in DNA synthesis was computed to be at around 01.00 hours (trough) by cosinor analysis, and correspondingly, the highest fraction of cells in DNA synthesis was found to be at around 13.00 hours (acrophase). A potential problem is that DNA synthesis may vary as a function of the site from which the BM sample is taken, both with regard to anatomical site and within the BM of the actual site.46Conceptually, one would regard the total red BM as one organ, being affected by the same endogenous physiologic and hormonal factors, and thus making site-dependent variations less important. This concept is supported by earlier reported data by Dosik et a1: who showed a very close correlation between DNA synthesis in BM samples obtained simultaneously from right and left iliac crests by biopsies. A good reproducibility, although with larger individual variations, was also demonstrated by bilateral simultaneous aspirations. In agreement with the results of Dosik et al;’we found no significant difference in S-phase between the right and left iliac crests. In the present study sampling was also performed from sternum, the BM from which showing a slightly higher S-phase. For obvious reasons it was not feasible to take samples from different localizations at the same time, and in our study protocol we tried to minimize this potential site-related problem by harvesting BM from different sites (ie, sternum and iliac crests) at the same timepoints for different individuals. The circadian stage-dependent rhythm was most pronounced for the sternum samples, but the same pattern of circadian variation was demonstrated for the samples from the iliac crests. For both localizations, the time of acrophase and trough was at about the same time, and accordingly corresponded to the acrophase and trough of the combined data. This finding rules out the possibility of /> d 00 04 08 12 16 20 20 Time of the day M.E.T.) different sampling sites contributing to the observed circadian stage dependence of DNA synthesis. Thus, the demonstration of nearly the same value of DNA synthesis in the left and right iliac crests indicates strongly that the total red BM must be looked on as a functional entity, and the same circadian variation in the BM of the sternum as of the iliac crests further corroborates this functional homogeneity. The higher DNA synthesis found in the sternum samples is most likely due to less blood contamination of these samples. This finding implies that the circadian stage-dependent variations demonstrated in this study may be even larger. It should be noticed that the DNA synthesis values found during daytime in the present study (14.3 f 0.4 [mean] and 15.6 f 0.5 [highest]) are not much different from DNA values obtained by trephine and from filtered BM fragments, 15.3% and 16.5%, respectively, reported by Zbroja et aLa This strongly suggests that the BM samples analyzed in our study are representative of BM nucleated cells. The major reason for this relatively high DNA synthesis in the aspirate may be the small amount of BM harvested for &w cytometry analysis. The finding that there was no statistically significant difference in cortisol and DNA synthesis measured 24 hours apart for the 19 24-hour periods contradicts the possibility of a stress-induced circadian rhythm of the DNA synthesis. In addition, a possible stress effect related to sampling was minimized by starting sampling at three different clock hours. These three factors, therefore, strongly negate the notion of a stress-induced circadian rhythm of the DNA synthesis. Likewise, it was found that neither the differential count of the BM smears nor a dilution effect of mature granulocytes leaving the BM compartment could explain the circadian variation in DNA synthesis observed (results not shown). Our data are in good agreement with those of Mauer, who found that [3H]TdR-labeled cells of the myeloid lineage were clearly higher during the day as compared with midnight in three of four individuals, and with a trend From www.bloodjournal.org by guest on June 16, 2017. For personal use only. 2610 SMAALAND ET AL towards lower DNA synthesis at midnight in the fourth individual. In that study, BM was sampled four times at different clock hours during a 42-hour time span, with 12 to 18 hours between time points.M Although the study of Mauer was an in vitro study because exposing individuals to the isotope was not feasible, the short exposure for 1 hour after the sample was obtained should express the DNA synthesis at the time of sampling. Killmann et al made corresponding observations in one human volunteer in an earlier study with regard to mitotic indices, demonstrating an increase in this proliferative parameter from early morning to late evening.29 The study reported here is thus a major expansion of these two earlier studies, and is in addition an in vivo study analyzing the proliferative activity at the exact time of sampling, demonstrating a statistically significant reproducible circadian stage-dependent variation in DNA synthesis of the human BM. The potential importance of the data presented here is underlined by the results recently published by U v i et al for murine BMT demonstrating a corresponding circadian variation in CFU-GM and DNA synthesis, with highest and lowest values of these parameters in the activity and rest span, respectively. The benefit of taking such rhythms into consideration was demonstrated in the same study by a circadian toxicity rhythm of the anticancer agent 4‘-0tetrahydropyranyl doxorubicin, demonstrating the lowest toxicity when the CFU-GM and DNA synthesis were lowest. Therefore, by taking circadian stage-dependent variations in DNA synthesis into account it may be possible to reduce BM toxicity of S-phase-specific drugs or drugs having a major effect on DNA synthesis by administering the drugs or the major dose of a continuous drug infusion during the time of lowest proliferative activity, ie, late evening or at night in diurnally active individuals. Cells in the S-phase will then be less susceptible, and cells in the GO/G1 phase will have more time for repairing damage before entering into the S-phase. However, in this context pharmacokinetic and pharmacodynamic properties of drugs must also be considered. An additional important aspect of these findings is that it may be possible to increase the effect of biologic response modifiers like GM-CSF and G-CSF by administering the optimal dose at the time of greatest responsiveness of the BM. This timed administration may increase their usefulness and effect, and possibly also reduce their side effects. The data further suggest that it may be possible to increase the fraction of proliferating cells with careful selection of time of day for harvesting BM cells for auto- or allografting. However, the existence of interindividual differences in circadian time structure, as well as the possibility that patients may have a different inherent time structure due to their illness, suggests that BM sampling should be performed at the two timepoints of anticipated highest and lowest DNA synthesis to verify this circadian pattern. This verification is easily performed by the sampling procedure described here, which is a routine clinical procedure. It is important that a small amount of BM aspirate is harvested to reduce the “contamination” of peripheral blood. Another way to individualize chronotherapy will be to monitor marker parameters in peripheral blood related to BM cell proliferative activity. Such studies are underway. Indeed, one may speculate that in the years to come, the optimization of doses, intervals and scheduling of doses, and treatment courses of both existing and upcoming drugs and biologic substances will become just as important as the institution of novel drugs and biologic response modifiers. The possibility of treating patients at an optimal circadian time, interval, and schedule is today feasible and costeffective through programmable drug delivery systems. ACKNOWLEDGMENT We are indebted to the volunteer subjects in the study. The authors gratefully acknowledge the skilful technical assistance of Jan Solsvik, Gro OlderQy, and Dagny Ann Sandnes. REFERENCES 1. Gale RP: Myelosuppressive effects of antineoplastic chemotherapy, in Testa NG, Gale RP (eds): Hematopoiesis. Long-Term Effects of Chemotherapy and Radiation. New York, NY, Marcell Dekker, 1988, p 63 2. Evans WE: Clinical pharmacodynamics of anticancer drugs: A basis for extending the concept of dose-intensity. Blut 56241, 1988 3. Hryniuk W, Figueredo A, Goodyear M: Applications of dose intensity to problems in chemotherapy of breast and colorectal cancer. Semin Oncol14:3, 1987 (suppl4) 4. Lohrman H-P, Schreml W: Cytotoxic Drugs and the Granulopoietic System. New York, NY,Springer-Verlag, 1982 5. Pollak MN: Recombinant GM-CSF in myelosuppression of chemotherapy. N Engl J Med 320:253,1989 (letter) 6. Haus E, Halberg F, Scheving LE, Pauly JE, Cardoso S, Kiihl JFW, Sothern RB, Shiotsuka RN, Hwang DS: Increased tolerance of leukemic mice to arabinosyl cytosine with schedule adjusted to circadian system. Science 177230,1972 7. Scheving LE, Haus E, Kiihl JFW, Pauly JE, Halberg F, Cardoso SS: Close reproduction by different laboratories of characteristics of circadian rhythm in 1-P-D-arabinofuranosykytosine tolerance by mice. Cancer Res 36:1133,1976 8. U v i F, Hrushesky WJM, Blomquist C, Lakatua D, Haus E, Halberg F, Kennedy BJ: Reduction of cisdiamminedichloroplatinum nephrotoxicity in rats by optimal circadian drug timing. Cancer Res 42:950,1982 9. Kiihl JFW, Haus E, Halberg F, Scheving LE, Pauly JE, Cardoso SS, Rosene G: Experimental chronotherapy with ara-C. Comparison of murine ara-C tolerance on differently timed treatment schedules. Chronobiologia 1:316,1974 10. Scheving LE, Bums ER, Halberg F, Pauly JE: Combined chronochemotherapy of L1210 leukemic mice using 1-P-D-arabinofuranosylcytosine, cyclophosphamide, vincristine, methylprednisolone, and cis-platinum. Chronobiologia 17:33,1980 11. Scheving LE, Burns ER, Pauly JE, Halberg F: Circadian bioperiodic response of mice bearing advanced L1210 leukemia to combination therapy with adriamycin and cyclophosphamide. Cancer Res 40:1511,1980 From www.bloodjournal.org by guest on June 16, 2017. For personal use only. BONE MARROW DNA SYNTHESIS RHYTHM 12. Sothern RB, E v i F, Haus E, Halberg F, Hrushesky WJM: Control of a murine plasmacytoma with doxorubicin-cisplatin: Dependence on circadian stage of treatment. J Natl Cancer Inst 81:135, 1989 13. von Roemeling R, Hrushesky WJM: Determination of the therapeutic index of floxuridine by its circadian infusion pattern. J Natl Cancer Inst 82386,1990 14. Scheving LE, Bums ER, Pauly JE, Halberg F, Haus E: Survival and cure of leukemic mice after circadian optimization of treatment with cyclophosphamide and 1-p-D-arabinofuranosykytosine. Cancer Res 373648,1977 15. Stoney PJ, Halberg F, Simpson HW: Circadian variation in colony-forming ability of presumably intact murine bone marrow cells. Chronobiologia 2:319,1975 16. Scheving LE, Burns ER, Pauly JE, Tsai TH: Circadian variation in cell division of the mouse alimentary tract, bone marrow, and corneal epithelium, and its possible implication in cell kinetics and cancer chemotherapy. Anat Rec 191:479,1978 17. Bartlett P, Haus E, Tuason T, Sackett-Lundeen L, Lakatua D: Circadian rhythm in number of erythroid and granulocytic colony forming units in culture (ECFU-C and GCFU-C) in bone marrow of BDFl male-mice, in Haus E, Kabat H F (eds): Proceedings of the XV International Conference of the International Society for Chronobiology. Basel, Switzerland, Karger, 1982, p 160 18. Aardal NP, Laerum OD: Circadian variations in mouse bone marrow. Exp Hematol 11:792,1983 19. Haus E, Lakatua DJ, Swoyer J, Sackett-Lundeen L Chronobiology in hematology and immunology. Am J Anat 168:467,1983 20. Aardal NP: Circannual variations of circadian periodicity in murine colony-forming cells (CFU-C). Exp Hematol12:61, 1984 21. Sletvold 0, Laerum OD: Multipotent stem cell (CFU-S) numbers and circadian variations in aging mice. Eur J Hematol 41:230,1988 22. Sletvold 0, Laerum OD, Riise T: Age-related differences and circadian and seasonal variations in myelopoietic progenitor cell (CFU-GM) numbers in mice. Eur J Haematol40:42,1988 23. Hrushesky WJM: Circadian timing of cancer chemotherapy. Science 228:73,1985 24. Kerr DJ, Lewis C, O’Neill B, Lawson N, Blackie RG, Newell DR, Boxall F, Cox J, Rankin EM, Kaye SB: The myelotoxicity of carboplatin is influenced by the time of its administration. Hematol Oncol8:59,1990 25. E v i F, Benavides M, Chevelle C, Le Saunier F, Bailleul F, Misset J-L, Regensberg C, Vannetzel J-M, Reinberg A, Mathe G: Chemotherapy of advanced ovarian cancer with 4’-O-tetrahydropyranyl doxorubicin and cisplatin: A randomized phase I1 trial with an evaluation of circadian timing and dose-intensity. J Clin Oncol 23705,1990 26. Rivard GE, Infante-Rivard C, Hoyoux C, Champagne J: Maintenance chemotherapy for childhood acute lymphoblastic leukemia: Better in the evening. Lancet 2:1264, 1985 27. Hrushesky WJM, von Roemeling R, Sothem RB: Circadian chronotherapy: From animal experiments to human cancer chemotherapy, in Lemmer B (ed): Chronopharmacology. Cellular and Biochemical Interactions. New York, NY, Marcel Dekker, 1989, p 439 28. Hrushesky WJM, von Roemeling R, Sothern RB: Preclinical and clinical cancer chemotherapy, in Arendt J, Minors DS, Waterhouse JM (eds): Biological Rhythms in Clinical Practice. London, UK, Wright, 1989, p 225 29. Killmann S-A, Cronkite EP, Fliedner TM, Bond VP: Mitotic indices of human bone marrow cells. I. Number and cytologic distribution of mitosis. Blood 19:743, 1962 2611 30. Mauer AM: Diumal variation of proliferative activity in the human bone marrow. Blood 26:1,1965 31. Ginsberg L, Ludman PF, Anderson JV, Burrin JM, Joplin GF: Does stressful venepuncture explain increased midnight serum cortisol concentration? Lancet 2:1257, 1988 32. VindelBv LL: Flow microfluorometric analysis of nuclear DNA in cells from solid tumors and cell suspensions. Virchows Arch [B] 24:227,1977 33. Dean PN, Jett JH: Mathematical analysis of DNA distributions derived from flow microfluorometry. J Cell Biol60523, 1974 34. Gray JW, Dean PN: Display and analysis of flow cytometric data. Annu Rev Biophys Bioeng 9509, 1980 35. Nelson W, Tong Y,Lee JK, Halberg F Methods for cosinor rhythmometry. Chronobiologia 6:305,1979 36. Van Cauter E: Endocrine rhythms, in Arendt J, Minors DS, Waterhouse JM (eds): Biological Rhythms in Clinical Practice. London, UK, Wright, 1989, p 23 37. Kleveu RB, Shymko RM, Blumenfeld D, Braly PS: Circadian gating of S phase in human ovarian cancer. Cancer Res 47:6267,1987 38. Hryniuk W, Bush H: The importance of dose intensity in chemotherapy of metastatic breast cancer. J Clin Oncol 2:1281, 1984 39. DeVita V Jr: Dose-response is alive and well (editorial). J Clin Oncol4:1157,6986 40. Simon R, Korn E L Selecting drug combinations based on total equivalent dose (dose intensity). J Natl Cancer Inst 82:1469, 1990 41. Moormeier JA, Williams SF, Kaminer LS, Garner M, Bitran JD: High-dose tri-alkylator chemotherapy with autologous stem cell rescue in patients with refractory malignancies. J Natl Cancer Inst 82:29, 1990 42. Cheson BD, Lacerna L, Leyland-Jones B, Sarosy G, Wittes RE: Autologous bone marrow transplantation. Ann Intern Med 11051, 1989 43. Bronchud MH, Scarfe JH, Thatcher N, Crowther D, Souza LM, Alton NK, Testa NG, Dexter TM: Phase 1/11 study of recombinant human granulocyte colony-stimulating factor in patients receiving intensive chemotherapy for small cell lung cancer. Br J Cancer 56:809,1988 44. Antman KS, Griffin JD, Elias A, Socinski MA, Yan L, Cannistra SA, Oette D, Whitley M, Frei E 111, Schnipper LE: Effect of recombinant human granulocyte-macrophage colony stimulating factor on chemotherapy-induced myelosuppression. N Engl J Med 319593,1988 45. McLachlin JR, Eglitis MA, Ueda K, Kantoff PW, Pastan IH, Anderson WF, Gottesman MM: Expression of a human complementary DNA for the multidrug resistance gene in murine hematopoietic precursor cells with the use of retroviral gene transfer. J Natl Cancer Inst 821260,1990 46. Lord BI: The architecture of bone marrow cell populations. Int J Cell Cloning 8:317,1990 47. Dosik GD, Barlogie B, Gohde W, Johnston D, Tekell JL, Drewinko B: Flow cytometry of DNA content in human bone marrow: A critical reappraisal. Blood 55:734,1980 48. Zbroja RA, Wass J, Vincent PC, Young GAR: Fragment filtration: A method for the accurate determination of flow cytometric kinetic data from bone marrow aspirates. Exp Hematol 14:85, 1986 49. Ltvi F, Blazsek I, Ferle-Vidovic A Circadian and seasonal rhythms in murine bone marrow colony-forming cells affect tolerance for the anticancer agent 4’-O-tetrahydropyranyladriamycin (THP). Exp Hematol16:696,1988 From www.bloodjournal.org by guest on June 16, 2017. For personal use only. 1991 77: 2603-2611 DNA synthesis in human bone marrow is circadian stage dependent R Smaaland, OD Laerum, K Lote, O Sletvold, RB Sothern and R Bjerknes Updated information and services can be found at: http://www.bloodjournal.org/content/77/12/2603.full.html Articles on similar topics can be found in the following Blood collections Information about reproducing this article in parts or in its entirety may be found online at: http://www.bloodjournal.org/site/misc/rights.xhtml#repub_requests Information about ordering reprints may be found online at: http://www.bloodjournal.org/site/misc/rights.xhtml#reprints Information about subscriptions and ASH membership may be found online at: http://www.bloodjournal.org/site/subscriptions/index.xhtml Blood (print ISSN 0006-4971, online ISSN 1528-0020), is published weekly by the American Society of Hematology, 2021 L St, NW, Suite 900, Washington DC 20036. Copyright 2011 by The American Society of Hematology; all rights reserved.
© Copyright 2024 Paperzz