目的：這個回朔性的研究目的在於分析未感染Epstein-Barr病毒之頭頸部原發部位不明癌症治療結果。材料和方法：自西元1993年3月至2002年十月期間，共有35位頭頸部原發部位不明癌症病人在林口長庚醫院放射腫瘤科接受治療，所有的病人均接受淋巴節或鼻咽腔的切片，其中有13位病人因為頸部淋巴結或鼻咽腔切片藏有EB病毒而被排除。總共22位經切片證實未感染EB病毒之頭頸部原發不明癌症病人被納入這次的分析。14 (63%)位病人接受手術併術後同步化學和放射線治療（7位）或術後放射線治療（7位）。其餘未接受手術的病人分別接受同步化學和放射線治療（3位）、放射線治療（3位）、或前導性化學治療及放射線治療（2位）。放射線治療的總劑量範圍從6000至7280 cGy，總劑量之中位數為6750 cGy。治療的單次劑量為180或200 cGy並且每個禮拜治療5天。結果：追蹤時間之中位數為8年（5-14年）。追蹤期間，有9病人仍然存活，13位死亡。9位病人發生局部復發或是治療後仍是無法控制的頸部腫塊，只有2位病人發生遠端轉移。3年、5年、和8年整體存活率分別為50%、50%和42%。以淋巴結分期為N1、N2和N3的病人其5年之整體存活率分別為100%、44%和33%（P值為0.019）。在多變數分析則僅有同時侵犯上半部和下半部頸部在統計學上有意義。3年和5年之區域控制率分別為60%和53%。在N1、N2和N3分期其5年之區域控制率分別為100%、67%和17%（P值為0.034）。在多變數分析則僅有同時侵犯上半部和下半部頸部在統計學上有意義。如果病人一開始即接受頸部擴清術，五年的整體存活率和區域控制率分別為64.3%和66.5%。相反地，如果一開始未接受手術，五年的整體存活率和區域控制率分別降到25%和31%。雖然未達到統計學上有意義，但仍可發現一開始就接受手術的病人在預後有較佳的結果。在淋巴結分期N3的病人同樣可以發現較佳的結果。結論：治療未感染EB病毒之頭頸部原發不明癌症上，在頸部淋巴結的區域控制是極富挑戰性的，在我們的研究中，這群病人若一開始未接受頸部擴清術或是同時侵犯上半部和下半部頸部其頸部的控制率是較差的。我們強調在頭頸部原發不明的癌症上頸部擴清術有其必要性，尤其在沒有EB病毒感染的病人。

Purpose: This retrospective study was aimed to analyze the treatment results of unknown primary head and neck cancer without Epstein-Barr virus (EBV) infection. Methods and Materials: During the period between March 1993 and October 2002, 35 patients with neck lymph node metastases from unknown primary head and neck cancers were treated at the Department of Radiation Oncology, Chang Gung Memorial Hospital, Linkou. Twenty-two of them were eligible for this study, because 13 patients were proven to have EBV infection in their nasopharynx or neck lymph nodes. Fourteen (64%) received surgery and postoperative concurrent chemoradiotherapy (7 patients) or radiotherapy alone (7 patients). Others who did not receive surgery were treated by concurrent chemoradiotherapy (3 patients), radiotherapy (3 patients), or induction chemotherapy with radiotherapy (2 patients). Total prescription dose ranged from 6000 cGy to 7280 cGy with a median of 6750 cGy to Waldeyer's ring and involved neck. Results: The median follow-up duration was 8 years (ranged from 5 to 14 years). Nine (41%) patients were alive, and 13 (59%) patients died. Nine (41%) patients had neck recurrence, including four patients with uncontrolled neck mass enlargement. Only 2 patients developed distant metastases. The 3-year, 5-year, and 8-year overall survival were 50%, 50%, and 42%, respectively. The 5-year overall survival for N1, N2, and N3 patients were 100%, 44%, and 33% (P value=0.019), respectively. In multivariant analysis, only both upper and lower neck involvement was a significant predictor for overall survival. The 3-year and 5-year regional control rates were 60% and 53%, respectively. For N1, N2, and N3 diseases, the 5-year regional control rates were 100%, 67%, and 17% (P value=0.034), respectively. In multivariant analysis, only both upper and lower neck involvement reached statistical significance for regional control. If patients initially were treated with neck node dissection, 5-year overall survival and regional control rate were 64% and 67%, respectively. If surgery was not performed initially, 5-year overall survival and regional control rate were 25% and 31%, respectively. Although all statistics did not reach significance, we still found a trend that favor surgery initially group. Better results were also noted in N3 patients. Conclusion: The treatment for neck nodal control of unknown primary head and neck cancer without EBV infection is challenging. In our present study, this group of patients had poor neck control if neck dissection was not performed initially or both upper and lower neck involvement. Neck dissection is suggested for patients with unknown primary head and neck cancer, especially those without EBV infection.