In{ J Radrnlion Oncology Biol Phys.. Vol. Printed in the U.S.A. All rights reserved.
18. pp.
1283-1286 Copyright
0360.3016/90 $3.00 + .Ml Q 1990 Pergamon Press plc
??Original Contribution
CELLULAR
RADIOSENSITIVITY CELL CARCINOMAS
OF PRIMARY AND LOCAL
HEAD AND NECK SQUAMOUS TUMOR CONTROL
WILLIAM A. BROCK, PH.D.,*
FRASER L. BAKER, PH.D.,* JENNIFER L. WIKE, B.S.,* SUSAN L. SIVON, B.S.* AND LESTER J. PETERS, M.D.+
The University of Texas M. D. Anderson Cancer Center, 15 15 Holcombe Blvd., Houston, TX 77030 The low dose survival parameters of human tumor cell lines have been shown to correlate with the perceived clinical radiosensitivity of different tumor histologic types. This conclusion has been generated from the analysis of a large number of cell lines and has, therefore, served as the basis for attempts to develop predictive assays of tumor radiocurability. In this study, the tumors from 72 patients with head and neck squamous cell carcinoma have been grown in an adhesive tumor cell assay system and their sensitivity to radiation has been measured. All patients in this study were treated with post-operative radiotherapy, the surgical margins were negative, and any patient that had received chemotherapy was excluded. The average S2 (survival at 2.0 Gy) value of the 72 cultures was 0.33, with the values ranging from 0.11 to 0.91. All patients were evaluated for local tumor control. They have been followed for about 1 year and continued follow-up is still in progress. The average survival at 2.0 Gy of cultures derived from the 12 patients that have had recurrences so far is slightly higher (0.40) than that from those who appear so far to have local tumor control (0.30). Although the general trend is that recurrent tumors yield primary cultures that are slightly more resistant, the difference is not statistically significant. Survival at 2.0 Gy, Primary culture, Predictive assays, Squamous cell carcinoma, Human.
This notion has been supported, in the case of radiation biology, by the analyses of Fertil and Malaise ( 10, 11) and Deacon et al. (9), who found that the radiosensitivity of
INTRODUCIION
goal of predictive assays is to measure one or more tumor characteristics that would identify those patients who are unlikely to be cured by radiotherapy so that they can then be considered for alternative or more aggressive therapy. It is well accepted that tumor growth kinetics, hypoxia, clonogenic cell number, or intrinsic tumor cell sensitivity to radiation all have potential for influencing tumor response to radiotherapy and are good candidates for predictive assays of tumor radiocurability ( 12). However, the relative importance of any parameter as a predictor will not be known until measurements have been made on individual patients and the results compared with treatment outcome. In this study, we are concentrating on the in vitro cellular radiosensitivity of head and neck squamous cell carcinomas. The rationale for measuring the in vitro sensitivity of cells to cytotoxic agents is based upon the stem cell model for human tumors (13), which hypothesizes that tumor stem cells are the target of therapy and that the in vitro sensitivity of the stem cells reflects their in vivo sensitivity. The
cell lines of tumor are more lines that
derived from human tumors was characteristic type and that, on the average, tumor types that difficult to cure by radiotherapy produce cell are more resistant to low doses of radiation. The survival curve parameter that they found best distinguished different tumor histologic types was survival at 2.0 Gy (S2) of radiation or any other descriptor of the low dose region of survival curves. More recently it has been shown in mice that in vitro and in vivo tumor cell killing are correlated (4, 5). Sufficient rationale, therefore, exists to justify making S2 measurements on primary cultures derived from human tumors and to compare the results with local tumor control by radiotherapy. In this study, we report the correlations between S2 measurements made in primary cultures derived from head and neck squamous cell carcinomas (SCC) and local tumor control, The 72 patients included in this study have been followed for over 1 year. Results from this study should reveal whether cellular ra-
research grants CA-50192, and CA-06294 from the National Cancer Institute, and by the Katharine Unsworth Memorial Fund. Accepted for publication 8 February 1990.
* Dept. of Experimental Radiotherapy. + Dept. of Clinical Radiotherapy. Reprint requests to: William A. Brock, Ph.D. Acknowledgements-We wish to thank Sheri Lee Axtell for manuscript preparation. This investigation was supported by 1283
1284
I. J. Radiation Oncology 0 Biology 0 Physics
diosensitivity plays a measurable role in local tumor control by radiotherapy. METHODS
AND MATERIALS
Culture methodology and survival assay All materials were the same as originally described (3) except as noted below. Precoated CamPlates were supplied.* Amphotericin B and gentamycin were obtained? and used at 1.25 5g/ml and 23 mg/ml, respectively, in addition to the other antibiotics; they are necessary to control contamination in head and neck tumor cultures, and tests have shown that they have no effect on cell growth at these concentrations. Cell attachment medium consists of serum-containing medium with the addition of 0.6% Methylcellulose 4000.$ Biopsy and surgical specimens were obtained from the Tissue Procurement Laboratory in the Department of Pathology at The University of Texas M. D. Anderson Cancer Center, where a pathologist dissects normal tissue free from tumor. Whenever necessary patients signed informed consent forms after the nature of the procedure was fully explained to them. All tissue was stored in complete medium at 40°C until enzyme disaggregation. Cell suspensions were performed as reported previously (1-3, 8), except that the digestion time was reduced to 5 hr. Single cell suspensions were inoculated in four 24-well CamPlates as reported previously (7). After 24 hr of incubation at 37°C the attachment medium was removed, and the cultures were washed with 1 ml of phosphate-buffered saline and fed 1 ml of growth medium. Rows of four wells were then irradiated from the bottom of each plate using a 250 kVp x-ray machine” fitted with a special lead shield that protected all other wells on the plate. The dose rate was 1.5 Gy/min, the maximum scatter to adjacent wells was less than 2%, and the dose range was 0.5 to 6.0 Gy. After 2 weeks of incubation with a medium change after 6 days, the medium was removed, all plates were rinsed with phosphate-buffered saline, fixed in 70% ethanol, air dried, and then stained with fresh 0.5% crystal violet in 20% ethanol. The integrated optical density of the stained wells was measured over the entire surface of each culture well with an image analysis system. Survival curves were calculated only from culture wells that were growing exponentially at the time of fixation. The ratio of optical densities of irradiated and control cultures was fitted to the linear quadratic model, and survival at 2.0 Gy was calculated from the parameters. Patients All patients had XC of the head and neck. They were treated with surgery and postoperative radiotherapy; the
* LifeTract, Inc.; Irvine, CA. + GIBCO; Grand Island, NY.
June 1990. Volume 18, Number 6 Table I. Head and neck tumor
specimens No.
Specimens received Low cell yield Cultures set up Technical errors Contamination Low growth Evaluable results
121 15 106 4 9 21 72
clinical correlations reported here are from these 72 postoperative radiotherapy patients. The patients were all assessed for risk of recurrence based upon pathological findings using criteria reported elsewhere; S2 values were not correlated with any pathological finding. Immunocytochemistry The epithelial origin of cells grown in each culture was assessed using an antibody cocktail composed of a mixture of monoclonal antibodies reactive against both acid and base cytokeratins (AE 1 and AE3) and three monoclonals directed against differentiation-specific surface antigens of epithelial tumor cells (260F9, 113F1, and 3 17G5) (14). The results of these assays with head and neck squamous cell carcinomas have been reported elsewhere (6).
RESULTS Culture characterization Head and neck squamous cell carcinoma cultures are tested to determine whether cell growth is of epithelial origin. The results of this method have been reported previously (6). Heterogeneity in the degree of staining intensity is common in these cultures, even if all cells show positive staining. A monoclonal antibody cocktail is necessary because a single antibody has not been found that will universally recognize epithelial specific antigens expressed by all head and neck tumors. Assay success rate Table 1 shows the ultimate fates of head and neck tumor specimens when they are processed for primary culture and radiation testing. During the period of this study, we received 12 1 squamous cell carcinoma surgical specimens and biopsies. Of these, 106 yielded enough cells for radiosensitivity testing. Of these 34 did not yield an evaluable result, mainly because of low growth. This resulted in 72 cases for this study or an overall success ratio of
* Fisher Scientific;
BPhillips.
Houston,
TX.
1285
Radiosensitivity and local tumor control 0 W. A. BROCK et al. Table 2. Variations
within assays, tumors,
Parameter
No.
CV(%)
Multiple assays on same cell suspension Multiple biopsies on the same patient S2 values of all SCC patients
21 I5 72
11.2 23.7 44.7
CV = Coefficient
DISCUSSION
and patients
variation.
(the percentage of cultures set up that gave an evaluable result).
68%
Variation in S2 measurements To test for intra-assay variation, multiple assays were performed on several tumor cell suspensions. In addition, survival measurements were made on up to five biopsy specimens from each patient to estimate interassay variation. Finally, the variation in head and neck SCC assays from 72 patients was calculated (Table 2). The relatively low intra-assay coefficient of variation (11.2%) is due to the fact that multiple assays were performed after the cell suspension had been made, thus making pipetting errors and differences in handling cultures the main sources of variation. However, multiple biopsies from the same tumor were treated independently throughout the entire procedure, so this source of variation (23.7%) includes differences in all steps of the procedure, including sampling differences of the tumor and intratumor heterogeneity. Finally, a coefficient of variation of 44.7% was calculated for all 72 head and neck SCC S2 values. It is the difference between this value and the intratumor variation that gives S2 potential as a predictive parameter. Clinical results All patients were treated by postoperative radiotherapy and are being followed to determine whether local tumor control was achieved. All patients had negative surgical margins, which means that many of them had been cured prior to radiotherapy; those who were not, only had microscopic disease at the time of radiotherapy. Therefore, this postsurgery group is composed only of patients who are more or less uniform with respect to tumor burden. Even though as many as 50% of these patients may have been cured by surgery alone, we know that the ones with recurrence were not and it should be meaningful to compare this group to the population as a whole. Even though this group has a disadvantage for analysis because of the complicating factor of surgery, they probably do not have gross differences in tumor cell numbers. The mean S2 value for cultures from the 12 patients whose disease recurred was 0.40, whereas the remaining population of 60 cultures had a mean of 0.30, compared to the entire group of 72 which had an average value of 0.33. Figure I shows cumulative frequency histograms of both groups. The differences in the groups is not statistically significant.
In this study we have concentrated on patients with SCC of the head and neck. We are attempting to determine if the in vitro sensitivity of tumor specimens correlates with the probability of local tumor control. At present, there does not appear to be a difference between any of the patient groups, but by continuing and expanding this study, differences may appear. The sensitivities of the cultures derived from these patients ranged from 0.11 to 0.9 1, with the majority of the values spanning the relatively small range of 0.21 to 0.42. When compared with our previous data or other data published from established human tumor cell lines, these cultures have about the same range of sensitivity. However, the mean value (S2 = 0.33) indicates head and neck SCC are more sensitive than several other tumor types, including melanoma, most adenocarcinomas, and SCC cultures derived from lung tumors (6). The clinical correlations for 72 patients who have been followed so far, show that cultures derived from recurrent tumors have an average S2 value slightly greater than the population that has not yet recurred. However, since there were only 12 recurrences in this group, more data must be accumulated. The S2 values of controlled and recurrent tumors overlap significantly, suggesting that other predictive parameters are needed for a useful predictive assay. One reason for the overlap might be the cellular heterogeneity of some tumors with respect to radiation sensitivity. Tofilon et al. (15) have recently published evidence that up to 20% of tumors contain more than one population published evidence that up to 20% of tumors contain more than one population of cells with different radiation sensitivities. The result of an in vitro S2 measurement on a heterogeneous primary culture would reflect the most sensitive population. Because tumors fail to be controlled by radiotherapy for more reasons than cellular resistance, it is unlikely that the results of any one assay will show a clear rela-
1
*-*recurrences
t0.1
0.2
Survival
mean 0.5
at 2.0
52
for all pts 1.0
Gy
Fig. 1. Cumulative frequency histograms of S2 values for all 72 patients in the study compared to only those with local recurrences. These data were updated as of September, 1989.
1286
1. J. Radiation Oncology 0 Biology 0 Physics
tionship with tumor control. However, even though the results of these studies may not yield a predictive assay of radiocurability, with a large patient population it should
June 1990, Volume 18, Number 6
be possible to determine whether a different S2 value is associated with tumors that are controlled with radiotherapy than those that are not.
REFERENCES 1. Ajani, J. A.; Baker, F. L.; Spitzer, G.; Kelly, A.; Brock, W. A.; Tomasovic, B.; Singletary, S. E.; McMurtrey, M.; Plager, C. Comparison between clinical response and in vitro drug sensitivity of primary human tumors in the adhesive tumor cell culture system. J. Clin. Oncol. 5: 19 12- 192 1; 1987. 2. Baker, F. L.; Ajani, J.; Spitzer, G.; Tomasovic. B. J.: Williams, M.; Finders, M.; Brock, W. A. High colony-forming efficiency of primary human tumor cells cultured in the adhesive-tumor-cell culture system: improvements with medium and serum alterations. Int. J. Cell Clon. 6:95-105: 1988. 3. Baker, F.; Spitzer, G.; Ajani, J. A.; Brock, W. A.; Lukeman, J.; Pathak, S.; Tomasovic, B.; Thielvoldt, D.; Williams, M.; Vines, C.; Tofilon, P. Drug and radiation sensitivity measurements of successful primary monolayer culturing of human tumor cells using cell-adhesive matrix and supplemented medium. Cancer Res. 46: 1263-1274: 1986. 4. Bristow, R. G.; Harley, P. A.; Hill, R. P. Comparisons beand in vivo radioresponse of tween in vim radiosensitivity murine tumour cell lines. I. Parameters of in vitro radiosensitivity and endogenous cellular glutathione levels. Int. J. Radiat. Oncol. Biol. Phys. 18: 133- 145; 1990. 5. Bristow, R. G.; Hill, R. P. Comparison between in vifro radiosensitivity and in vivo radioresponse of murine tumor cell lines II: in vivo radioresponse following fractionated treatment and in vitro/in vivo correlations. Int. J. Radiat. Oncol. Biol. Phys. 18:33 l-345; 1990. 6. Brock, W. A.; Baker, F. L.; Peters, L. J. Radiosensitivity of human head and neck squamous cell carcinomas in primary culture and its potential as a predictive assay of tumor radiocurability. Int. J. Radiat. Biol. 56:75 l-760; 1989. 7. Brock, W. A.; Maor, M. H.; Peters, L. J. Predictors oftumor
8.
9.
IO.
I I.
12.
13.
14.
15.
response to radiotherapy. Radiat. Res. 8(Suppl):S290-S296; 1985. Brock, W. A.; Williams, M.; Bhadkamkar, V. A.; Spitzer, G.; Baker, F. Radiosensitivity testing of primary cultures derived from human tumors. In: Karcher, K. H., Kogelnik, H. D., Szepesi, T., ed. Progress in radio-oncology III. Vienna: International Club for Radio-Oncology/Austria; 1987:300306. Deacon, J.; Peckham, M. J.; Steel, G. G. The radioresponsiveness of human tumours and the initial slope of the cell survival curve. Radiother. Oncol. 2:3 17-323; 1984. Fertil, B.; Malaise, E. P. Inherent cellular radiosensitivity as a basic concept for human tumor radiotherapy. Int. J. Radiat. Oncol. Biol. Phys. 7:62 l-629; 198 1. Fertil, B.; Malaise, E. P. Intrinsic radiosensitivity of human cell lines is correlated with radioresponsiveness of human tumors: analysis of 101 published survival curves. Int. J. Radiat. Oncol. Biol. Phys. 11:1699-1707; 1985. Peters, L. J.; Brock, W. A.; Chapman, J. D.; Wilson, G. Predictive assays of tumor radiocurability. Amer. J. Clin. Oncol. 1I :275-287; 1988. Selby, P.; Buick, R. N.; Tannock, I. A critical appraisal of the “human tumor stem-cell assay.” N. Engl. J. Med. 308: 129-134: 1983. Taha, M.; Ordone, N. G.; Kulkarni, S.; Owen, M.; Hortobagyi, G.; Reading, C. L.; Dicke, K. A.; Spitzer, G. A monoclonal antibody cocktail for detection of micrometastatic tumor cells in bone marrow. Bone marrow transplant. 4: 297-303; 1989. Tofilon, P. J.; Vines, C. M.; Meyn, R. E.; Wike, J.; Brock, W. A. Heterogeneity in radiation sensitivity within human primary tumor cell cultures as detected by the SCE assay. Br. J. Cancer 59:54-60; 1989.
18. pp.
1283-1286 Copyright
0360.3016/90 $3.00 + .Ml Q 1990 Pergamon Press plc
??Original Contribution
CELLULAR
RADIOSENSITIVITY CELL CARCINOMAS
OF PRIMARY AND LOCAL
HEAD AND NECK SQUAMOUS TUMOR CONTROL
WILLIAM A. BROCK, PH.D.,*
FRASER L. BAKER, PH.D.,* JENNIFER L. WIKE, B.S.,* SUSAN L. SIVON, B.S.* AND LESTER J. PETERS, M.D.+
The University of Texas M. D. Anderson Cancer Center, 15 15 Holcombe Blvd., Houston, TX 77030 The low dose survival parameters of human tumor cell lines have been shown to correlate with the perceived clinical radiosensitivity of different tumor histologic types. This conclusion has been generated from the analysis of a large number of cell lines and has, therefore, served as the basis for attempts to develop predictive assays of tumor radiocurability. In this study, the tumors from 72 patients with head and neck squamous cell carcinoma have been grown in an adhesive tumor cell assay system and their sensitivity to radiation has been measured. All patients in this study were treated with post-operative radiotherapy, the surgical margins were negative, and any patient that had received chemotherapy was excluded. The average S2 (survival at 2.0 Gy) value of the 72 cultures was 0.33, with the values ranging from 0.11 to 0.91. All patients were evaluated for local tumor control. They have been followed for about 1 year and continued follow-up is still in progress. The average survival at 2.0 Gy of cultures derived from the 12 patients that have had recurrences so far is slightly higher (0.40) than that from those who appear so far to have local tumor control (0.30). Although the general trend is that recurrent tumors yield primary cultures that are slightly more resistant, the difference is not statistically significant. Survival at 2.0 Gy, Primary culture, Predictive assays, Squamous cell carcinoma, Human.
This notion has been supported, in the case of radiation biology, by the analyses of Fertil and Malaise ( 10, 11) and Deacon et al. (9), who found that the radiosensitivity of
INTRODUCIION
goal of predictive assays is to measure one or more tumor characteristics that would identify those patients who are unlikely to be cured by radiotherapy so that they can then be considered for alternative or more aggressive therapy. It is well accepted that tumor growth kinetics, hypoxia, clonogenic cell number, or intrinsic tumor cell sensitivity to radiation all have potential for influencing tumor response to radiotherapy and are good candidates for predictive assays of tumor radiocurability ( 12). However, the relative importance of any parameter as a predictor will not be known until measurements have been made on individual patients and the results compared with treatment outcome. In this study, we are concentrating on the in vitro cellular radiosensitivity of head and neck squamous cell carcinomas. The rationale for measuring the in vitro sensitivity of cells to cytotoxic agents is based upon the stem cell model for human tumors (13), which hypothesizes that tumor stem cells are the target of therapy and that the in vitro sensitivity of the stem cells reflects their in vivo sensitivity. The
cell lines of tumor are more lines that
derived from human tumors was characteristic type and that, on the average, tumor types that difficult to cure by radiotherapy produce cell are more resistant to low doses of radiation. The survival curve parameter that they found best distinguished different tumor histologic types was survival at 2.0 Gy (S2) of radiation or any other descriptor of the low dose region of survival curves. More recently it has been shown in mice that in vitro and in vivo tumor cell killing are correlated (4, 5). Sufficient rationale, therefore, exists to justify making S2 measurements on primary cultures derived from human tumors and to compare the results with local tumor control by radiotherapy. In this study, we report the correlations between S2 measurements made in primary cultures derived from head and neck squamous cell carcinomas (SCC) and local tumor control, The 72 patients included in this study have been followed for over 1 year. Results from this study should reveal whether cellular ra-
research grants CA-50192, and CA-06294 from the National Cancer Institute, and by the Katharine Unsworth Memorial Fund. Accepted for publication 8 February 1990.
* Dept. of Experimental Radiotherapy. + Dept. of Clinical Radiotherapy. Reprint requests to: William A. Brock, Ph.D. Acknowledgements-We wish to thank Sheri Lee Axtell for manuscript preparation. This investigation was supported by 1283
1284
I. J. Radiation Oncology 0 Biology 0 Physics
diosensitivity plays a measurable role in local tumor control by radiotherapy. METHODS
AND MATERIALS
Culture methodology and survival assay All materials were the same as originally described (3) except as noted below. Precoated CamPlates were supplied.* Amphotericin B and gentamycin were obtained? and used at 1.25 5g/ml and 23 mg/ml, respectively, in addition to the other antibiotics; they are necessary to control contamination in head and neck tumor cultures, and tests have shown that they have no effect on cell growth at these concentrations. Cell attachment medium consists of serum-containing medium with the addition of 0.6% Methylcellulose 4000.$ Biopsy and surgical specimens were obtained from the Tissue Procurement Laboratory in the Department of Pathology at The University of Texas M. D. Anderson Cancer Center, where a pathologist dissects normal tissue free from tumor. Whenever necessary patients signed informed consent forms after the nature of the procedure was fully explained to them. All tissue was stored in complete medium at 40°C until enzyme disaggregation. Cell suspensions were performed as reported previously (1-3, 8), except that the digestion time was reduced to 5 hr. Single cell suspensions were inoculated in four 24-well CamPlates as reported previously (7). After 24 hr of incubation at 37°C the attachment medium was removed, and the cultures were washed with 1 ml of phosphate-buffered saline and fed 1 ml of growth medium. Rows of four wells were then irradiated from the bottom of each plate using a 250 kVp x-ray machine” fitted with a special lead shield that protected all other wells on the plate. The dose rate was 1.5 Gy/min, the maximum scatter to adjacent wells was less than 2%, and the dose range was 0.5 to 6.0 Gy. After 2 weeks of incubation with a medium change after 6 days, the medium was removed, all plates were rinsed with phosphate-buffered saline, fixed in 70% ethanol, air dried, and then stained with fresh 0.5% crystal violet in 20% ethanol. The integrated optical density of the stained wells was measured over the entire surface of each culture well with an image analysis system. Survival curves were calculated only from culture wells that were growing exponentially at the time of fixation. The ratio of optical densities of irradiated and control cultures was fitted to the linear quadratic model, and survival at 2.0 Gy was calculated from the parameters. Patients All patients had XC of the head and neck. They were treated with surgery and postoperative radiotherapy; the
* LifeTract, Inc.; Irvine, CA. + GIBCO; Grand Island, NY.
June 1990. Volume 18, Number 6 Table I. Head and neck tumor
specimens No.
Specimens received Low cell yield Cultures set up Technical errors Contamination Low growth Evaluable results
121 15 106 4 9 21 72
clinical correlations reported here are from these 72 postoperative radiotherapy patients. The patients were all assessed for risk of recurrence based upon pathological findings using criteria reported elsewhere; S2 values were not correlated with any pathological finding. Immunocytochemistry The epithelial origin of cells grown in each culture was assessed using an antibody cocktail composed of a mixture of monoclonal antibodies reactive against both acid and base cytokeratins (AE 1 and AE3) and three monoclonals directed against differentiation-specific surface antigens of epithelial tumor cells (260F9, 113F1, and 3 17G5) (14). The results of these assays with head and neck squamous cell carcinomas have been reported elsewhere (6).
RESULTS Culture characterization Head and neck squamous cell carcinoma cultures are tested to determine whether cell growth is of epithelial origin. The results of this method have been reported previously (6). Heterogeneity in the degree of staining intensity is common in these cultures, even if all cells show positive staining. A monoclonal antibody cocktail is necessary because a single antibody has not been found that will universally recognize epithelial specific antigens expressed by all head and neck tumors. Assay success rate Table 1 shows the ultimate fates of head and neck tumor specimens when they are processed for primary culture and radiation testing. During the period of this study, we received 12 1 squamous cell carcinoma surgical specimens and biopsies. Of these, 106 yielded enough cells for radiosensitivity testing. Of these 34 did not yield an evaluable result, mainly because of low growth. This resulted in 72 cases for this study or an overall success ratio of
* Fisher Scientific;
BPhillips.
Houston,
TX.
1285
Radiosensitivity and local tumor control 0 W. A. BROCK et al. Table 2. Variations
within assays, tumors,
Parameter
No.
CV(%)
Multiple assays on same cell suspension Multiple biopsies on the same patient S2 values of all SCC patients
21 I5 72
11.2 23.7 44.7
CV = Coefficient
DISCUSSION
and patients
variation.
(the percentage of cultures set up that gave an evaluable result).
68%
Variation in S2 measurements To test for intra-assay variation, multiple assays were performed on several tumor cell suspensions. In addition, survival measurements were made on up to five biopsy specimens from each patient to estimate interassay variation. Finally, the variation in head and neck SCC assays from 72 patients was calculated (Table 2). The relatively low intra-assay coefficient of variation (11.2%) is due to the fact that multiple assays were performed after the cell suspension had been made, thus making pipetting errors and differences in handling cultures the main sources of variation. However, multiple biopsies from the same tumor were treated independently throughout the entire procedure, so this source of variation (23.7%) includes differences in all steps of the procedure, including sampling differences of the tumor and intratumor heterogeneity. Finally, a coefficient of variation of 44.7% was calculated for all 72 head and neck SCC S2 values. It is the difference between this value and the intratumor variation that gives S2 potential as a predictive parameter. Clinical results All patients were treated by postoperative radiotherapy and are being followed to determine whether local tumor control was achieved. All patients had negative surgical margins, which means that many of them had been cured prior to radiotherapy; those who were not, only had microscopic disease at the time of radiotherapy. Therefore, this postsurgery group is composed only of patients who are more or less uniform with respect to tumor burden. Even though as many as 50% of these patients may have been cured by surgery alone, we know that the ones with recurrence were not and it should be meaningful to compare this group to the population as a whole. Even though this group has a disadvantage for analysis because of the complicating factor of surgery, they probably do not have gross differences in tumor cell numbers. The mean S2 value for cultures from the 12 patients whose disease recurred was 0.40, whereas the remaining population of 60 cultures had a mean of 0.30, compared to the entire group of 72 which had an average value of 0.33. Figure I shows cumulative frequency histograms of both groups. The differences in the groups is not statistically significant.
In this study we have concentrated on patients with SCC of the head and neck. We are attempting to determine if the in vitro sensitivity of tumor specimens correlates with the probability of local tumor control. At present, there does not appear to be a difference between any of the patient groups, but by continuing and expanding this study, differences may appear. The sensitivities of the cultures derived from these patients ranged from 0.11 to 0.9 1, with the majority of the values spanning the relatively small range of 0.21 to 0.42. When compared with our previous data or other data published from established human tumor cell lines, these cultures have about the same range of sensitivity. However, the mean value (S2 = 0.33) indicates head and neck SCC are more sensitive than several other tumor types, including melanoma, most adenocarcinomas, and SCC cultures derived from lung tumors (6). The clinical correlations for 72 patients who have been followed so far, show that cultures derived from recurrent tumors have an average S2 value slightly greater than the population that has not yet recurred. However, since there were only 12 recurrences in this group, more data must be accumulated. The S2 values of controlled and recurrent tumors overlap significantly, suggesting that other predictive parameters are needed for a useful predictive assay. One reason for the overlap might be the cellular heterogeneity of some tumors with respect to radiation sensitivity. Tofilon et al. (15) have recently published evidence that up to 20% of tumors contain more than one population published evidence that up to 20% of tumors contain more than one population of cells with different radiation sensitivities. The result of an in vitro S2 measurement on a heterogeneous primary culture would reflect the most sensitive population. Because tumors fail to be controlled by radiotherapy for more reasons than cellular resistance, it is unlikely that the results of any one assay will show a clear rela-
1
*-*recurrences
t0.1
0.2
Survival
mean 0.5
at 2.0
52
for all pts 1.0
Gy
Fig. 1. Cumulative frequency histograms of S2 values for all 72 patients in the study compared to only those with local recurrences. These data were updated as of September, 1989.
1286
1. J. Radiation Oncology 0 Biology 0 Physics
tionship with tumor control. However, even though the results of these studies may not yield a predictive assay of radiocurability, with a large patient population it should
June 1990, Volume 18, Number 6
be possible to determine whether a different S2 value is associated with tumors that are controlled with radiotherapy than those that are not.
REFERENCES 1. Ajani, J. A.; Baker, F. L.; Spitzer, G.; Kelly, A.; Brock, W. A.; Tomasovic, B.; Singletary, S. E.; McMurtrey, M.; Plager, C. Comparison between clinical response and in vitro drug sensitivity of primary human tumors in the adhesive tumor cell culture system. J. Clin. Oncol. 5: 19 12- 192 1; 1987. 2. Baker, F. L.; Ajani, J.; Spitzer, G.; Tomasovic. B. J.: Williams, M.; Finders, M.; Brock, W. A. High colony-forming efficiency of primary human tumor cells cultured in the adhesive-tumor-cell culture system: improvements with medium and serum alterations. Int. J. Cell Clon. 6:95-105: 1988. 3. Baker, F.; Spitzer, G.; Ajani, J. A.; Brock, W. A.; Lukeman, J.; Pathak, S.; Tomasovic, B.; Thielvoldt, D.; Williams, M.; Vines, C.; Tofilon, P. Drug and radiation sensitivity measurements of successful primary monolayer culturing of human tumor cells using cell-adhesive matrix and supplemented medium. Cancer Res. 46: 1263-1274: 1986. 4. Bristow, R. G.; Harley, P. A.; Hill, R. P. Comparisons beand in vivo radioresponse of tween in vim radiosensitivity murine tumour cell lines. I. Parameters of in vitro radiosensitivity and endogenous cellular glutathione levels. Int. J. Radiat. Oncol. Biol. Phys. 18: 133- 145; 1990. 5. Bristow, R. G.; Hill, R. P. Comparison between in vifro radiosensitivity and in vivo radioresponse of murine tumor cell lines II: in vivo radioresponse following fractionated treatment and in vitro/in vivo correlations. Int. J. Radiat. Oncol. Biol. Phys. 18:33 l-345; 1990. 6. Brock, W. A.; Baker, F. L.; Peters, L. J. Radiosensitivity of human head and neck squamous cell carcinomas in primary culture and its potential as a predictive assay of tumor radiocurability. Int. J. Radiat. Biol. 56:75 l-760; 1989. 7. Brock, W. A.; Maor, M. H.; Peters, L. J. Predictors oftumor
8.
9.
IO.
I I.
12.
13.
14.
15.
response to radiotherapy. Radiat. Res. 8(Suppl):S290-S296; 1985. Brock, W. A.; Williams, M.; Bhadkamkar, V. A.; Spitzer, G.; Baker, F. Radiosensitivity testing of primary cultures derived from human tumors. In: Karcher, K. H., Kogelnik, H. D., Szepesi, T., ed. Progress in radio-oncology III. Vienna: International Club for Radio-Oncology/Austria; 1987:300306. Deacon, J.; Peckham, M. J.; Steel, G. G. The radioresponsiveness of human tumours and the initial slope of the cell survival curve. Radiother. Oncol. 2:3 17-323; 1984. Fertil, B.; Malaise, E. P. Inherent cellular radiosensitivity as a basic concept for human tumor radiotherapy. Int. J. Radiat. Oncol. Biol. Phys. 7:62 l-629; 198 1. Fertil, B.; Malaise, E. P. Intrinsic radiosensitivity of human cell lines is correlated with radioresponsiveness of human tumors: analysis of 101 published survival curves. Int. J. Radiat. Oncol. Biol. Phys. 11:1699-1707; 1985. Peters, L. J.; Brock, W. A.; Chapman, J. D.; Wilson, G. Predictive assays of tumor radiocurability. Amer. J. Clin. Oncol. 1I :275-287; 1988. Selby, P.; Buick, R. N.; Tannock, I. A critical appraisal of the “human tumor stem-cell assay.” N. Engl. J. Med. 308: 129-134: 1983. Taha, M.; Ordone, N. G.; Kulkarni, S.; Owen, M.; Hortobagyi, G.; Reading, C. L.; Dicke, K. A.; Spitzer, G. A monoclonal antibody cocktail for detection of micrometastatic tumor cells in bone marrow. Bone marrow transplant. 4: 297-303; 1989. Tofilon, P. J.; Vines, C. M.; Meyn, R. E.; Wike, J.; Brock, W. A. Heterogeneity in radiation sensitivity within human primary tumor cell cultures as detected by the SCE assay. Br. J. Cancer 59:54-60; 1989.
