Girinsky T, Bernheim A, Lubin R, Tavakoli-Razavi T, Baker F, Janot F, Wibault P, Cosset J M, Duvillard P, Duverger A
Department of Radiation Oncology, CNRS UA 1158, Villejuif, France.
Int J Radiat Oncol Biol Phys. 1994 Nov 15;30(4):789-94. doi: 10.1016/0360-3016(94)90350-6.
To determine whether in vivo parameters (surviving fraction at 2 Gy, alpha values, and calculated cell growth fraction) were predictive of the treatment outcome.
Biopsies were obtained from patients with a head and neck tumor. In vitro parameters were determined using the CAM plate assay. Cell characterization by cytogenetic analysis was performed on 19 different cell cultures. In 25 additional cell cultures, cell clonogenicity was tested using the Courtenay Mills assay.
Biopsies were obtained from 156 patients with a head and neck tumor and the oropharynx was the predominant primary site. In vitro parameters were obtained in 113 cases (72%) (SF2 in 93 cases and calculated cell growth fraction in 103 cases). Cell characterization showed that cells in CAM plates were diploid with no clonal chromosome abnormalities and gave colonies in soft agar with a mean cloning efficiency of 1.610(-3). Only patients treated with surgery and/or radiation (76), were considered eligible for in vitro parameters and treatment outcome correlation studies. The mean follow-up is over 2 years (range 9-47 months). The local control rate was significantly higher (p = 0.04) for patients with alpha values above the cut-off point of 0.07 Gy-1 (69% vs. 38% at 2 years). The local control rate was also significantly higher (p = 0.04) for patients with calculated cell growth fraction values about the cut-off point of 0.06% (70% vs. 48% at 2 years). Moreover for these latter patients the overall survival rate was also significantly higher (p = 0.004) (54% vs. 26% at 2 years). It is worth noting that alpha and calculated cell growth fraction values below the cut-off points identified a small group of patients (about 20%) who were at a significantly high risk of local failure. From a pragmatic point of view, as only radiosensitivity or calculated cell growth fraction values could be obtained in a certain number of experiments due to technical reasons, the treatment outcome of patients who had either alpha and/or calculated cell growth fraction values below the cut-off levels (about 30% of all patients) was analyzed. This group of patients fared significantly worse (p = 0.02) in terms of local control (50% vs. 68% at 2 years) and (p = 0.04) overall survival (36% vs. 50% at 2 years).
These results suggest that in vitro parameters using the CAM plate assay, might be useful in predicting the treatment outcome of patients with a head and neck tumor treated with surgery and postoperative radiation, or radiation alone. However, they must be considered as preliminary because the cut offs used in the study were chosen for exploratory purposes. Only a multivariate analysis including all clinical and biologic factors will allow us to draw any firm conclusions.
确定体内参数（2 Gy时的存活分数、α值和计算得出的细胞生长分数）是否可预测治疗结果。
从头颈肿瘤患者身上获取活检样本。使用克隆形成能力分析方法（CAM平板法）测定体外参数。对19种不同的细胞培养物进行细胞遗传学分析以进行细胞特征鉴定。在另外25种细胞培养物中，使用考特尼·米尔斯法检测细胞克隆形成能力。
从156名头颈肿瘤患者身上获取了活检样本，口咽是主要的原发部位。113例（72%）患者获得了体外参数（93例有2 Gy时的存活分数，103例有计算得出的细胞生长分数）。细胞特征表明，CAM平板中的细胞为二倍体，无克隆性染色体异常，且在软琼脂中形成集落，平均克隆效率为1.6×10⁻³。仅接受手术和/或放疗的患者（76例）被认为符合进行体外参数与治疗结果相关性研究的条件。平均随访时间超过2年（范围9 - 47个月）。α值高于0.07 Gy⁻¹临界值的患者局部控制率显著更高（p = 0.04）（2年时为69%对38%）。计算得出的细胞生长分数值高于0.06%临界值的患者局部控制率也显著更高（p = 0.04）（2年时为70%对48%）。此外，对于后一组患者，总生存率也显著更高（p = 0.004）（2年时为54%对26%）。值得注意的是，α值和计算得出的细胞生长分数值低于临界值的患者识别出一小部分（约20%）局部失败风险显著较高的患者。从实际角度来看，由于技术原因，在一定数量的实验中只能获得放射敏感性或计算得出的细胞生长分数值，因此对α值和/或计算得出的细胞生长分数值低于临界水平的患者（约占所有患者的30%）的治疗结果进行了分析。这组患者在局部控制方面（2年时为50%对68%）（p = 0.02）和总生存率方面（2年时为36%对50%）（p = 0.04）明显更差。
这些结果表明，使用克隆形成能力分析方法（CAM平板法）获得的体外参数可能有助于预测接受手术和术后放疗或单纯放疗的头颈肿瘤患者的治疗结果。然而，这些结果必须被视为初步的，因为研究中使用的临界值是为探索目的而选择的。只有包括所有临床和生物学因素在内的多变量分析才能使我们得出任何确凿的结论。
