Brief Report Serum Thymidine Kinse 1 Activity in the Prognosis nd Monitoring of Chemotherpy in Lung Cncer Ptients: A Brief Report Benjmin Nismn, PhD,* Hovv Nechushtn, MD, PhD,* Him Birn, MD, Hds Gntz-Sorotsky, MD,* Nir Peled, MD, Simon Gronowitz, PhD, nd Tmr Peretz, MD* Introduction: Thymidine kinse 1 (TK1) is metbolic enzyme involved in DNA synthesis. Most stndrd tretment protocols for lung cncer (LC) include cytotoxic gents, which re potentil modultors of TK1. We imed to ssess the prognostic significnce of serum TK1 ctivity nd its role in monitoring chemotherpy in LC ptients. Methods: TK1 ctivity ws mesured using the DiviTum (Biovic) ssy in ser from 233 ptients with non smll-cell lung cncer (NSCLC), 91 with smll-cell lung cncer (SCLC), nd 90 with benign lung disese. Results: TK1 ctivity ws significntly ssocited with ge, performnce sttus, nd stge in NSCLC nd with stge nd weight loss in SCLC. In multivrite nlysis, pretretment TK1 ctivity, djusted for performnce sttus, stge, nd weight loss, independently ffected survivl in NSCLC (reltive risk =1.45, p = 0.031) nd SCLC (reltive risk = 2.49, p = 0.001). In NSCLC ptients, djusted elevted TK1 ctivity (>100 Du/L) t pretretment ws significnt predictor of tretment filure (odds rtio = 2.55, p = 0.01). A smll (less thn twofold) increse in TK1 ctivity fter the first nd second cycle of chemotherpy ws significntly ssocited with tretment filure nd poor overll survivl. Conclusions: Elevted pretretment serum TK1 ctivity ws n independent, dverse prognostic fctor, bsed on survivl, in the two min histologicl types of LC. A smll (less thn twofold) increse in TK1 ctivity fter the first nd second cycle of chemotherpy ws ssocited with tretment filure nd poor overll survivl. Key Words: Thymidine kinse 1, NSCLC, SCLC, Prognosis, Monitoring chemotherpy (J Thorc Oncol. 2014;9: 1568 1572) *Deprtment of Oncology, Hdssh nd Hebrew University Medicl Centre, Jeruslem, Isrel; Lung Cncer Unit, Division of Oncology, Sheb Medicl Center, Tel Hshomer, Isrel; nd Group of Clinicl Virology, Deprtment of Medicl Sciences, Uppsl University, Sweden. Disclosure: S.G. is the inventor of the TK1 ssy by DiviTum principle nd the founder of Biovic Interntionl AB, where his fmily members hold shres. The other uthors stte tht there re no conflicts of interest regrding the publiction of this rticle. Address for correspondence: Benjmin Nismn, PhD, Deprtment of Oncology, Hdssh nd Hebrew University Medicl Centre, P.O. Box 12000, Jeruslem 91120, Isrel. E-mil: nismn@hdssh.org.il DOI: 10.1097/JTO.0000000000000276 Copyright 2014 by the Interntionl Assocition for the Study of Lung Cncer ISSN: 1556-0864/14/0910-1568 One of the most importnt biologicl mechnisms of cncer ggressiveness is relted to uncontrolled tumor prolifertion; its mrkers hve been shown to hve prognostic vlue when mesured in lung cncer (LC) ptients. 1,2 To tke these mesurements, however, biopsy or resection my be required. One such mrker is thymidine kinse 1 (TK1), which is the key cytosolic enzyme in the slvge pthwy for deoxythymidine monophosphte (dtmp) synthesis; becuse TK1 is involved in DNA synthesis, it is considered mrker for cell prolifertion. Notbly, dividing cells relese TK1 during mitotic exit, which is medited by the ubiquitin system. 3 Thus, TK1 ctivity cn be detected in the serum. Recently, new highly sensitive ssy for the mesurement of TK ctivity in serum hs been developed tht is bsed on immobiliztion of the phosphorylted rection product by incorportion into DNA. 4 Korkmz et l. 5 evluted this ssy in smll group of 48 ptients with dvnced non smll-cell lung cncer (NSCLC) nd found significnt correltion between serum TK1 ctivity nd primry tumor mximum stndrdized uptke, bsed on 18-fluorodeoxyglucose positron emission tomogrphy scns, nd overll survivl. The present study ws imed to ssess the prognostic significnce of the new TK1 ssy in the two min histologicl types of LC, NSCLC nd smll-cell lung cncer (SCLC). Although TK1 levels detected in the serum primrily reflect prolifertive ctivity, this enzyme is n importnt element of complex system of kinses in the slvge pthwy, complementing the min de novo pthwy of dtmp synthesis. The ctivities of the two pthwys re coordinted with the trnsport systems of nucleosides. 6 There re mny fctors modifying these systems. Most stndrd tretment protocols for LC include cytotoxic gents tht re potentil modultors of TK1 ctivity. 6 8 Therefore, we investigted the usefulness of this ssy to monitor tretment with these gents. METHODS AND PATIENTS This prospective study ws performed on consecutive ptients between November 1, 2008, nd October 30, 2013. It ws pproved by the Institutionl Ethicl Review Bord (0441-08-HMO) nd included 90 ptients with benign lung disese 1568 Journl of Thorcic Oncology Volume 9, Number 10, October 2014
Journl of Thorcic Oncology Volume 9, Number 10, October 2014 Serum Thymidine Kinse 1 Activity in Lung Cncer TABLE 1. Distribution of TK1 Activity in Ptients With Benign Lung Disese or Lung Cncer Serum TK1 ctivity (Du/L) Study group n Men Medin IQR 95th Percentile p vlue All LC vs. BLD p <0.001 BLD 90 188 74 29 186 855 All LC 324 819 155 59 479 3481 All non smll-cell lung crcinom 233 468 129 56 344 1,484 Age (yr) p = 0.001 62 112 509 178 76 514 1,593 >62 121 424 101 36 241 1,307 Gender p = 0.259 Mle 151 578 148 60 397 1,705 femle 82 266 111 42 303 1,135 Smoking hbits p = 0.561 Nonsmoker 70 929 131 54 410 2, 184 Smoker/ex-smoker 163 332 129 58 289 1, 496 Histologicl type p = 0.571 Adenocrcinom 145 536 108 46 341 1,450 Squmous 59 393 181 60 358 1,520 Other 29 283 135 42 336 1,509 Performnce sttus p = 0.024 0 1 190 302 114 56 292 1,348 2 43 1216 222 82 697 3,138 Weight loss (kg) p = 0.469 5 94 518 115 58 325 1, 407 >5 139 400 150 55 373 2, 083 TNM stge p < 0.001 I II 39 153 82 31 152 1, 158 III 66 234 116 57 224 1, 213 IV 128 653 181 63 514 1, 711 All smll-cell lung crcinom 91 1,717 325 85 1478 6,132 Age p = 0.278 62 44 1, 986 423 102 1555 5, 947 >62 47 1, 465 195 64 1477 9, 795 Gender p = 0.118 Mle 66 2,148 398 95 2,273 10,086 femle 25 577 115 76 697 4, 036 Performnce sttus p = 0.177 0 1 54 1, 276 181 60 1,127 6, 104 2 37 2, 298 414 98 1,883 14,586 Weight loss (kg) p = 0.013 5 56 1, 018 159 50 843 5, 629 >5 35 2, 828 467 126 3,090 19,195 Stge p = 0.007 Limited 32 362 139 82 400 2, 160 Extensive 59 2, 452 535 95 3,090 12,124 IQR, interqurtile rnge; TK1, thymidine kinse 1; BLD, benign lung disese; LC, lung cncer. p vlues derived from Kruskl-Wllis nd Mnn-Whitney tests. (BLD), 233 ptients with NSCLC, nd 91 ptients with SCLC. The dignoses of ll lung tumors were confirmed pthologiclly. After informed consent ws given, serum smples were obtined from ll ptients, before the strt of tretment, nd stored t 80 C until nlysis. Serum TK1 ctivity ws mesured with colorimetric enzyme-linked immunosorbent ssy kit (DiviTum; Biovic Interntionl AB, Uppsl, Sweden), s described previously. 9 TK1 levels were expressed in DiviTum units/l (Du/L). Copyright 2014 by the Interntionl Assocition for the Study of Lung Cncer 1569
Nismn et l. Journl of Thorcic Oncology Volume 9, Number 10, October 2014 FIGure. 1. Survivl for 233 ptients with non smll-cell lung cncer (A) nd 91 ptients with smll-cell lung cncer (B). Kpln-Meier estimtes ccording to pretretment serum TK1 ctivity. TK1, thymidine kinse. Two tumor response ctegories were pplied: progressive disese (PD) nd nonprogression (NP), including complete remission, prtil response, nd stble disese. Sttisticl Methods Kruskl-Wllis, Mnn-Whitney, Wilcoxon, Spermn s correltion, Fisher s exct test, logistic regression, Kpln- Meier, nd Cox survivl models, nd receiver operting chrcteristic curves were used for sttisticl nlyses. RESULTS TK1 Activity Assy in NSCLC Ptients with both NSCLC nd SCLC showed significntly higher TK1 ctivities compred with those with BLD (medins 129 nd 325 versus 74 Du/L; p = 0.003 nd p < 0.001, respectively). Furthermore, ptients with SCLC hd higher TK1 ctivity thn those with NSCLC (p = 0.017). A cutoff vlue between low- nd high-serum TK1 ws set t 100 Du/L the closest point on the receiver operting chrcteristic curve to the optiml sensitivity nd specificity in the LC (NSCLC + SCLC) versus BLD model. In NSCLC (Tble 1), the serum TK1 ctivity ws ssocited with ge, performnce sttus (PS), nd disese stge. Medin survivl for ptients with norml versus elevted TK1 ctivity ws 18.6 versus 9.0 months, respectively (p = 0.001, Fig. 1A). Multivrite nlysis (Tble 2) selected PS, disese stge, weight loss (WL), nd serum TK1 s the most importnt pretretment chrcteristics with independent impcts on survivl. Of 233 ptients with NSCLC, 199 were treted with chemotherpy. The pretretment ctivity of TK1 in PD ptients (n = 74) ws significntly higher thn tht in NP ptients (n = 125) (medin 183 versus 105 Du/L, respectively, p = 0.005). In the logistic regression, the elevted TK1 ctivity (>100 Du/L), djusted for stge, PS, nd WL, ws significnt predictor of tretment filure (odds rtio [OR] = 2.55, 95% confidence intervl, 1.25 5.22). TK1 Activity Assy in SCLC In SCLC, TK1 levels were ssocited (Tble 1) with disese stge nd WL. The medin survivl for ptients with norml TK1 ctivity ws longer thn tht for those with elevted levels (16.5 versus 8.3 months, respectively, p = 0.003, Fig. 1B). TABLE 2. Multivrite Cox s Regression: Reltive Risk in Ptients with NSCLC (n = 233) or SCLC (n = 91) Chrcteristics Reltive risk 95% CI p vlue NSCLC Performnce sttus (0 1 vs. 2) 3.53 2.35 5.29 < 0.001 Weight loss (no vs. yes) 1.63 1.16 2.28 0.005 Stge (M0 vs. M1) 2.14 1.53 2.98 < 0.001 TK1 ( 100 Du/L vs. > 100 Du/L) 1.45 1.04 2.02 0.031 SCLC Performnce sttus (0 1 vs. 2) 1.83 1.11 3.02 0.018 Weight loss (no vs. yes) 1.77 1.07 2.92 0.026 Stge (limited vs. > extensive) 1.39 0.80 2.43 0.241 TK1 ( 100 Du/L vs. > 100 Du/L) 2.49 1.41 4.41 0.001 CI, confidence intervl; TK1, thymidine kinse 1; NSCLC, non smll-cell lung cncer; SCLC, smll-cell lung cncer. In multivrite nlysis, PS, WL, nd serum TK1 independently ffected survivl (Tble 2). Eighty-five of 91 ptients were treted with chemotherpy. The bsl TK1 ctivity in PD ptients (n = 21, medin 730 Du/L) ws higher thn tht in NP ptients (n = 64, medin 197 Du/L, p = 0.041). The elevted TK1 ctivity did not rech sttisticl significnce for the prediction of tretment filure (OR = 1.95, p = 0.09). TK1 Activity Assy in Monitoring Chemotherpy Ptients receiving pltinum-bsed therpy, combined with pemetrexed (n = 43), gemcitbine (n = 13), vinorelbine (n = 22), nd etoposide (n = 18), were monitored before the first (bsl), second, nd third cycles of chemotherpy. A significnt increse in TK1 ctivity ws observed fter the first nd second cycles for ll pplied regimens (Tble 3). However, the increse ws lrger for pemetrexed nd gemcitbine (medins, second/first rtio: 6.3 nd 8.1, respectively; medins, third/ first rtio: 12.4 nd 22.8, respectively) compred with vinorelbine nd etoposide (medins, second/first rtio: 1.5 nd 2.2, respectively; medins, third/first rtio: 1.8 nd 1.9, respectively). Ptients were divided into two groups: (1) treted with pemetrexed or gemcitbine nd (2) treted with vinorelbine or etoposide. TK1 ctivity significntly incresed in PD nd 1570 Copyright 2014 by the Interntionl Assocition for the Study of Lung Cncer
Journl of Thorcic Oncology Volume 9, Number 10, October 2014 Serum Thymidine Kinse 1 Activity in Lung Cncer TABLE 3. Chnges in TK1 Activity After the First nd Second Course of Chemotherpy, According to Regimen Pltinum-bsed regimen with Pemetrexed Gemcitbine Vinorelbine Etoposide Histology NSCLC NSCLC NSCLC SCLC Number 43 13 22 18 2nd/1st rtio (medin, IQR) p vlue (2nd vs. 1st) 3rd/1st rtio (medin, IQR) p vlue (3rd vs. 1st) 6.3, 3.6 20.6 12.4, 5.0 38.6 8.1, 4.7 30.3 0.003 22.8, 6.2 53.3 0.008 1.5, 1.1 2.4 1.8, 1.2 3.4 0.026 2.2, 1.2 11.0 0.008 1.9, 1.1 14.0 TK1, thymidine kinse 1; 1st, 2nd, 3rd, TK1 ctivity mesured before first (bsl), second, nd third cycles of chemotherpy, respectively; IQR, interqurtile rnge; NSCLC, non smll-cell lung cncer; SCLC, smll-cell lung cncer. p vlues were derived from the Wilcoxon signed-rnk test. TABLE 4. Chnges in TK1 Activity After the First nd Second Cycles of Chemotherpy According to Response Pltinum/pemetrexed or gemcitbine Pltinum/vinorelbine or etoposide Response PD NP PD NP Number 11 45 16 24 2nd/1st rtio (medin, IQR) p vlue (2nd vs. 1st) 3rd/1st rtio (medin, IQR) p vlue (3rd vs. 1st) 2.1, 1.3 15.4 0.003 1.9, 1.2 20.4 6.1, 1.4 15.4 6.9, 2.8 36.5 2.1, 1.5 4.7 0.030 1.5, 1.2 1.7 0.043 1.7, 1.17 10.3 0.004 3.0, 1.3 14.3 0.010 TK1, thymidine kinse 1; 1st, 2nd, 3rd, TK1 ctivity mesured before first (bsl), second, nd third cycles of chemotherpy, respectively; PD, progressive disese; NP, nonprogressive; IQR, interqurtile rnge. p vlues were derived from the Wilcoxon signed-rnk test. NP ptients in both groups fter the first nd second cycles of chemotherpy (Tble 4). However, this increse ws smller in PD ptients thn in NP ptients, with the exception of one subset of ptients in the NP group fter the first cycle of pltinum with vinorelbine or etoposide. After the second cycle, the rtio in this NP group reched 3. The study popultion ws split in two groups ccording to the increse in TK1 ctivity, using cutoff of 2.0 (overll medin increse in PD ptients for ll regimens). The increse ws more thn twofold fter the first cycle of chemotherpy in 12 of 27 PD ptients (44%) nd in 49 of 69 NP ptients (71%) (p = 0.021). After the second cycle, more thn twofold increse occurred in eight of 23 PD ptients (38%) nd 58 of 73 NP ptients (80%) (p = 0.003). Ptients with more thn twofold increse in TK1 ctivity fter the first (Fig. 2A) nd second (Fig. 2B) cycles of chemotherpy showed significntly longer medin survivl compred with those with less thn or equl to twofold increse (17.5 versus 9.9 months, p = 0.023 nd 19.6 versus 8.3 months, p = 0.002, respectively). We nlyzed reltionship between TK1 ctivity nd WBC ccount during chemotherpy. Spermn s correltion coefficients t regimens contining pemetrexed, gemcitbine, vinorelbine, nd etoposide were 0.13 (p = 0.202), 0.26 (p = 0.09), 0.43 (p = 0.003), nd 0.45 (p = 0.007), respectively. DISCUSSION The pretretment serum TK1 ctivity detected in this study reflects the overll condition (prolifertive bckground) of ech individul. The TK1 ctivity of the lung tumor nd its contribution to mesured TK1 ctivity in serum remin unknown. The higher TK1 ctivity of LC ptients versus tht of BLD is conceivbly due to the higher prolifertion of LC. The significnt ssocitions of TK1 ctivity with tumor burden both in NSCLC nd in SCLC my indicte, t lest in prt, its tumor origin. The DiviTum ssy explored in this study demonstrted the clinicl importnce of mesuring TK1 ctivity s prognostic mrker in NSCLC nd SCLC. In both subtypes of LC, elevted TK1 predicted poor survivl fter the djustment for disese stge, WL, nd PS. The present study is consistent with previous, retrospective study on TK1 ctivity, mesured by rdioenzyme immunossy. 10 The uthors of tht study pointed to the promising prognostic vlue of pretherpeutic TK for overll survivl in multivrite nlyses of NSCLC ptients. Moreover, we showed tht elevted pretretment TK1 ctivity in NSCLC, fter djustment for stge, PS, nd WL, remins significnt predictor of tretment filure with n OR of 2.55. We find tht pltinum-bsed combintions with pemetrexed, gemcitbine, vinorelbine, nd etoposide induced significnt increse in TK1 ctivity. However, the strongest increse ws observed in regimens tht included pemetrexed or gemcitbine, wheres less pronounced effect ws observed in those tht included etoposide or vinorelbine. Topolcn et l. 11 previously observed n increse in TK1 ctivity in colorectl cncer during chemotherpy with regimens tht included 5-fluorourcil (5-FU). Cytotoxic gents, such s pemetrexed, gemcitbine (through its metbolite, difluorodeoxyuridine monophosphte), nd 5-FU, re strong inhibitors of thymidylte synthse nd the de novo pthwy of dtmp synthesis. 6,12,13 To overcome this effect, cells up-regulte the compensting slvge pthwy for dtmp synthesis. The incresed ctivity of the slvge pthwy leds to n incresed uptke of thymidine Copyright 2014 by the Interntionl Assocition for the Study of Lung Cncer 1571
Nismn et l. Journl of Thorcic Oncology Volume 9, Number 10, October 2014 FIGure. 2. Kpln-Meier estimtes ccording to the rtio between two mesurements of TK1 ctivity before the first nd second (A) or third (B) cycles of chemotherpy in 96 ptients with lung cncer. TK1, thymidine kinse. through ctivtion of the membrne equilibrtive nucleoside trnsporter 1 (ENT1) delivery system. 6,14,15 Recent experimentl dt obtined with HeL cells tht highly express ENT1 demonstrted tht etoposide ws strong inducer of ENT1 ctivity nd TK1 expression. 8 Activtion of the slvge TK pthwy nd/or the ENT1 system ppers to be the min fctor explining incresed TK1 ctivity fter tretment with pemetrexed-, nd gemcitbine-, or etoposide-contining regimens. At these conditions, TK1 behves s metbolic mrker, reflecting chnges in two pthwys of dtmp synthesis. The increse of serum TK1 ctivity during chemotherpy my result from metbolic chnges both in tumor nd norml tissues. This is confirmed by the observtion of incresed TK1 ctivity t the use of 5-FU-contining regimens in djuvnt setting in colon cncer ptients. 11 The overll increse of TK1 ctivity might include for exmple the repliction effect from leukocyte. Higher increse denotes higher recovery of the leukocytes nd might contribute better tretment response to tretment. This mechnism might be importnt t the use of vinorelbine nd etoposide. Actully, t these regimens, we observed better correltion between the TK1 ctivity nd WBC counts during chemotherpy (r = 0.43 0.45). We found tht the extent of increse in TK1 ctivity t the beginning of LC tretment my predict its effectiveness nd prognosis. An nlysis of chnges during chemotherpy showed tht more thn twofold increse in enzyme ctivity ws more often found in NP ptients thn in PD ptients fter the first nd second cycles of chemotherpy. Survivl nlysis lso showed longer survivl for ptients who hd more thn twofold increse in TK1 ctivity fter either the first or the second cycle of chemotherpy. Thus, monitoring by TK1 erly during chemotherpy could provide vluble informtion for detection of nonresponders, thereby preventing unnecessry toxicity. Our results show tht the difference between two response groups ws evident lredy fter the first cycle of tretment with regimens contining thymidylte synthse blocking gents: pemetrexed nd gemcitbine. Other mechnisms pper to be responsible for the TK1 ctivity chnge t the use of etoposide/ vinorelbine, nd this my be reson for delying increse, which hs become evident only fter the second cycle. Further studies re needed to revel nd describe dditionl fctors contributing to overll TK1 condition during chemotherpy nd improving ssocitive models. REFERENCES 1. Hnhn D, Weinberg RA. The hllmrks of cncer. Cell 2000;100:57 70. 2. Tinnemns MM, Lenders MH, ten Velde GP, et l. Evlution of prolifertion prmeters in vivo bromodeoxyuridine lbelled lung cncers. Virchows Arch 1995;427:295 301. 3. Ke PY, Chng ZF. Mitotic degrdtion of humn thymidine kinse 1 is dependent on the nphse-promoting complex/cyclosome-cdh1-medited pthwy. Mol Cell Biol 2004;24:514 526. 4. Gronowitz JS. A method nd kit for determintion of thymidine kinse nd use thereof. Europen ptent EP1856275 B1, ptent filed Feb 24, 2006, certificte 12.08.09. 5. Korkmz T, Seber S, Okutur K, et l. Serum thymidine kinse 1 levels correltes with FDG uptke nd prognosis in ptients with non smll cell lung cncer. Biomrkers 2013;18:88 94. 6. Presscco J, Mitrovski B, Erlichmn C, Hedley DW. Effects of thymidylte synthse inhibition on thymidine kinse ctivity nd nucleoside trnsporter expression. Cncer Res 1995;55:1505 1508. 7. Longley DB, Hrkin DP, Johnston PG. 5-fluorourcil: mechnisms of ction nd clinicl strtegies. Nt Rev Cncer 2003;3:330 338. 8. Lee EJ, Lee SJ. Etoposide increses equilibrtive nucleoside trnsporter 1 ctivity nd fluorothymidine uptke: screening of 60 cytotoxic gents. Oncol Rep 2013;29:763 770. 9. Nismn B, Allweis T, Kdouri L,et l. Comprison of dignostic nd prognostic performnce of two ssys mesuring thymidine kinse 1 ctivity in serum of brest cncer ptients. Clin Chem Lb Med 2013;51: 439 447. 10. Holdenrieder S, Von Pwel J, Duell T, et l. Clinicl relevnce of thymidine kinse for the dignosis, therpy monitoring nd prognosis of nonoperble lung cncer. Anticncer Res 2010;30:1855 1862. 11. Topolcn O, Holubec L Jr, Finek J, et l. Chnges of thymidine kinse (TK) during djuvnt nd pllitive chemotherpy. Anticncer Res 2005;25:1831 1833. 12. Schultz RM, Ptel VF, Worzll JF, Shih C. Role of thymidylte synthse in the ntitumor ctivity of the multitrgeted ntifolte, LY231514. Anticncer Res 1999;19:437 443. 13. Mini E, Nobili S, Ccigli B, Lndini I, Mzzei T. Cellulr phrmcology of gemcitbine. Ann Oncol 2006;17(Suppl 5):v7 12. 14. Komori S, Osd S, Mori R, et l. Contribution of thymidylte synthse to gemcitbine therpy for dvnced pncretic cncer. Pncres. 2010;39:1284 1292. 15. Sprtlin JL, Mckey JR. Humn equilibrtive nucleoside trnsporter 1 (hent1) in pncretic denocrcinom: towrds individulized tretment decisions. Cncers (Bsel) 2010;2:2044 2054. 1572 Copyright 2014 by the Interntionl Assocition for the Study of Lung Cncer