Combined surgery and postoperative radiotherapy for cervical lymph node metastases from an unknown primary tumour

Combined surgery and postoperative radiotherapy for cervical lymph node metastases from an unknown primary tumour

Radiotherapy and Oncology 49 (1998) 33–40 Combined surgery and postoperative radiotherapy for cervical lymph node metastases from an unknown primary ...

151KB Sizes 0 Downloads 103 Views

Radiotherapy and Oncology 49 (1998) 33–40

Combined surgery and postoperative radiotherapy for cervical lymph node metastases from an unknown primary tumour Primozˇ Strojan a ,*, Aleksandar Anicˇin b a

Institute of Oncology, Zalosˇka 2, SI-1105 Ljubljana, Slovenia University Department of Otorhinolaryngology and Cervicofacial Surgery, Zalosˇka 4, SI-1105 Ljubljana, Slovenia

b

Received 22 December 1997; revised version received 1 June 1998; accepted 9 July 1998

Abstract Purpose: During the 20-year period under study, 125 squamous cell carcinoma cases were detected among 234 patients with cervical lymph node metastases from an unknown primary tumour diagnosed in Slovenia. Fifty-eight patients were treated by surgery and postoperative radiotherapy, 56 of whom were eligible for follow-up and treatment evaluation. Patients and methods: There were six females and 50 males ranging in age from 33 to 81 years (median 56 years). Neck dissection was performed in 48 patients, while eight patients underwent extirpation of a single lymph node metastasis. On histopathological examination, disease was assessed as pN1 in six patients, pN2 in 37 patients and pN3 in 13 patients. The degree of tumour differentiation was G1 in one patient, G2 in 26 patients, G3 in 22 patients and GX in seven patients. Extracapsular tumour spread was found in 37 out of 48 examined specimens. Postoperatively, the irradiation field covered different sites of possible occult primary in 48 patients and in eight patients it was limited to the involved side of the neck. The median tumour doses were 59 and 55 Gy, respectively. Results: After a median follow-up of 8.6 years, a total of five (9%) primary tumours were subsequently discovered, all in the head and neck region. Twenty-three (41%) patients were alive without evidence of disease and 14 (25%) patients died due to cancer-unrelated causes. Disease-related deaths occurred in 19 (34%) patients, the cause of death being primary tumour in three patients, advanced nodal disease in 10 patients and distant metastases with no evidence of persistent tumour in the head and neck region in six patients. The 5- and 10-year disease-specific survival rates were 66% and 52%, respectively, and the overall survival rates were 52% and 22%, respectively. The patients’ survival significantly correlated (P , 0.05) with extracapsular tumour spread and the extent of the irradiation field. Conclusions: With acceptable toxicity, a superior control rate of neck disease and survival results favourably comparable to those reported elsewhere and obtained by a single modality approach, our study supports the use of combined therapy. In patients with a poor clinical and histopathological profile a more aggressive treatment approach exploiting chemotherapy seems to be justified.  1998 Elsevier Science Ireland Ltd. All rights reserved Keywords: Cervical lymph node metastases; Unknown primary tumour; Squamous cell carcinoma; Combined therapy; Radiation therapy; Surgery

1. Introduction Metastatic cancer in neck lymph nodes from an unknown primary tumour represents 2–3% of the total head and neck cancer cases [17]. Histologically, it is most often squamous cell carcinoma. While it is usually the case for the upper and mid-neck masses to originate from the primary hidden somewhere in the head and neck region, those in the

* Corresponding author.

lower neck often arise from cancer below the clavicles and histological types other than squamous cell carcinoma should be considered [17,24]. Combined modality treatment (surgical removal of neck metastases and postoperative irradiation of the neck node regions alone or together with the potential primary tumour sites in the head and neck) represents a recommended treatment strategy in cases with squamous cell histology [2,6–8, 10,12,14–16,19,27]. Only for a single mobile node less than 3 cm in size or when pN1 stage of disease with no extracapsular extension is found on histopathological examina-

0167-8140/98/$19.00  1998 Elsevier Science Ireland Ltd. All rights reserved PII S0167-8140 (98 )0 0082-6

34

P. Strojan, A. Anicˇin / Radiotherapy and Oncology 49 (1998) 33–40

tion performed after initial surgery is combined therapy unnecessary and a comprehensive neck dissection or radiotherapy alone may be sufficient [6–8,10,12,15,17]. The reported overall neck failure in a series utilizing combined modality treatment is 13–26% and when at least part of the head and neck mucosa is irradiated electively, the subsequent detection of the primary varies from 6% to 16% [2,7,14,16,19,27]. The latter does not significantly differ from the incidence of metachronous head and neck cancer following definitive irradiation for a known primary lesion [10]. However, long-term survival remains poor, ranging from 35% to 66% at 5 years and from 19% to 59% at 10 years [2,7,14,16,19,27]. Moreover, a substantial proportion of patients are not able to undergo a radical treatment procedure. They are only treated with palliative intent, usually with various adapted radiotherapy regimens. This report presents our experience with combined surgery and postoperative radiotherapy in patients with metastatic squamous cell carcinoma of cervical lymph nodes from an unknown primary tumour. We analyzed the treatment results and survival of patients who were seen at the University Department of Otorhinolaryngology and Cervicofacial Surgery and at the Institute of Oncology in Ljubljana, Slovenia, between 1975 and 1994.

The routine diagnostic work-up comprised clinical examination and endoscopy of the aerodigestive tract, chest Xray and standard haematological and biochemical tests; after 1980, neck and abdominal ultrasonography were performed routinely. Further investigations (fiberbronchoscopy, gastroscopy, computed tomography, etc.) were done selectively in accordance with clinical indications. Biopsies of the suspect mucosal sites above the clavicles, most often from the nasopharynx and tonsils, were performed in 40 (71%) patients. The malignant nature of the neck lump was confirmed before surgery in 48 (86%) patients (by fine needle biopsy in 33 patients and by incisional or excisional biopsy in 15 patients). In the remaining eight patients, malignancy was diagnosed by frozen section on surgery. The clinical and tumour characteristics of the patients are listed in Table 1. The UICC TNM classification system was used to asses the anatomical extent of the disease [11]. The topographic distribution of metastatic lymph nodes, based on the classification by Shah et al. [23], was submandibular group in 14 patients, upper jugular in 39 patients, middle jugular in 19 patients, lower jugular in eight patients and posterior triangle group in nine patients (Fig. 1). 2.2. Treatment

2. Patients and methods 2.1. Patients In a 20-year period under study, 234 patients with cervical lymph node metastases, no other evident disease elsewhere in the body and no history of previous malignancy were registered in the 2 million population of Slovenia by the national Cancer Registry [4]. The histological type was squamous cell carcinoma in 125 (53%) patients, adenocarcinoma in 36 (15%) patients, anaplastic carcinoma in 14 (6%) patients, carcinoma not otherwise specified in 46 (20%) patients, lymphoma in five (2%) patients, melanoma in four (2%) patients and other in four (2%) patients. Of the patients who presented with squamous cell carcinoma, 58/ 125 (46%) were treated with a combination of surgery and postoperative radiotherapy, 50 (40%) had radiotherapy as a single treatment modality, seven (6%) underwent surgical resection alone, two (2%) had surgery combined with both pre- and postoperative radiotherapy and eight (6%) received only symptomatic medical treatment. Hereby we reported on 56/58 squamous cell carcinoma patients who were primarily treated with surgery and were postoperatively irradiated. Two patients from this group were excluded from the analysis because they were lost to follow-up immediately after completed therapy. There were six females and 50 males ranging in age from 33 to 81 years (median 56 years). The interval between the time when the neck lump was observed and the diagnosis ranged from 0.5 to 36 months (median 3 months).

All patients underwent surgery and postoperative radiotherapy. Neck dissection was performed in 48 (86%) patients and was extended to neighbouring structures (parotid gland, mandible and external carotid artery) in six patients. The extent of the neck dissection, according to the classification of Robbins et al. [22], was radical neck Table 1 Patient data and tumour parameters Characteristics Sex Male Female Age (years) (median (range)) pN stage N1 N2 N3 Involved lymph node regions Isolated Multiple Histopathological grade G1 G2 G3 GX Extracapsular spread Negative Positive Unknown

No. of patients (n = 56) (%) 50 (89) 6 (11) 56 (33–81) 6 (11) 37 (66) 13 (23) 33 (59) 23 (41) 1 26 22 7

(2) (46) (39) (13)

11 (20) 37 (66) 8 (14)

P. Strojan, A. Anicˇin / Radiotherapy and Oncology 49 (1998) 33–40

Fig. 1. Topographic distribution of metastatic lymph nodes based on the classification by Shah et al. [23]. The numbers in the cycles indicate the number of patients with metastatic lymph nodes.

dissection in 29 patients, modified radical neck dissection in seven patients, selective neck dissection in six patients and extended neck dissection in six patients. In eight patients a solitary lymph node metastasis was extirpated. Resection was complete in 45 (80%) patients, while 11 patients presented with macroscopic evidence of residual tumour after surgery. The time interval between surgery and the beginning of postoperative radiotherapy was 9–48 days (median 20 days). Most patients, i.e. 48 (86%), were irradiated on a Cobalt60 unit through two parallel opposite lateral fields to the nodes as well as to potential primary tumour sites in the head and neck (unknown primary tumour treatment volume). Whether the whole pharyngeal axis or only a part of it was included in the irradiation field depended on the level of nodal involvement and the patient’s life-style. For example, in those presenting with mid-cervical nodes or in heavy drinkers and/or smokers, the nasopharynx was spared. On the other hand, in younger patients, when there was a finding of jugulodigastric or posterior cervical nodes and visual or radiographic evidence of an irregularity of the nasopharynx with negative biopsy, the nasopharynx and oropharynx were routinely treated, sparing the hypopharynx and larynx. A tumour dose of 18–62 Gy (median 50 Gy) was administered in 1.8–2 Gy daily fractions applied five times weekly. A tumour dose of less than 50 Gy was delivered in 6/48 patients; radiotherapy was omitted on the patient’s request in five cases and due to unexpected

35

cancer-unrelated death of the patient in one case. Because of severe acute side-effects, a split-course regimen was used in five patients, with a 2–3 week free interval between the first (38–44 Gy) and second (16–24 Gy) irradiation course, to a tumour dose of 60–62 Gy. The rest of the patients (77%) received a tumour dose in the range of 50–54 Gy. Thereafter, the irradiation was limited to the area of the involved neck nodes which were boosted, depending on pathohistological characteristics of resected lymph nodes and the absence/presence of residual disease after surgery, up to a cumulative dose of 80 Gy (median 59 Gy) using a linear accelerator 8–12 MeV electron beam. After 40 Gy, the initial irradiation field was reduced for spinal cord protection. The lower neck and supraclavicular regions were treated by photons through a single anterior field to a tumour dose of 50 Gy delivered in 25 fractions of 2 Gy. Eight patients were irradiated by the modified technique because of poor histopathological prognostic factors (four patients), or metastases in the lower jugular and/or supraclavicular nodes (four patients). The irradiation field encompassed the involved side of the neck together with the ipsilateral supraclavicular region (the neck-unilateral technique); a tumour dose of 34–70 Gy (median 55 Gy) was delivered in daily fractions of 2–3 Gy each using a Cobalt60 unit photon beam and/or a linear accelerator 8–18 MeV electron beam. 2.3. Statistical methods Statistical calculations were performed using a PC computer and a Biomedical statistical software package (BMDP) [3]. The differences in the distribution of patients according to various clinical or histopathological characteristics were tested using contingency-table analysis and x2 testing with Yates’ adjustment, or the Fischer exact test when the number of patients was less than five in any cell of the table [1]. The primary end-point of our analysis was disease-specific survival (DSS), which was calculated by censoring deaths from diseases of unrelated causes. When calculating overall survival (OS), all deaths were considered as events. In both cases, the survival time was calculated from the date of diagnosis. Statistical assessment of patient survival was made using the Kaplan–Meier product limit method [13] and log-rank comparison to evaluate the differences between the survival curves [21]. A small number of patients did not allow the use of multivariate analysis. All tests were two-sided and statistical significance was calculated at the 5% level.

3. Results The median follow-up time was 8.6 years (range 1.6– 17.8 years); 79% of patients were followed for at least 5 years.

36

P. Strojan, A. Anicˇin / Radiotherapy and Oncology 49 (1998) 33–40

Fig. 2. Disease-specific survival (DSS) and overall survival (OS).

3.1. Survival DSS at 5 and 10 years was 66% (95% CI 52–79%) and 52% (95% CI 31–72%), respectively, and OS was 52% (95% CI 38–65%) and 22% (95% CI 5–38%), respectively (Fig. 2). A statistically significant difference in 5-year DSS rates was obtained when stratifying patients according to extracapsular spread and the extent of the irradiation field (Table 2 and Fig. 3). Other histopathological characteristics (i.e. pN stage of disease, the number of involved lymph node regions and grade), the type of surgical resection, the presence of residual tumour after surgery and the irradiaTable 2 Disease-specific survival (DSS) at 5 years: results of univariate analysis Variables

pN stage N1 N2 N3 Involved lymph node regions Isolated Multiple Histopathological grade G1 + G2 G3 Extracapsular extension Negative Positive Type of surgery Extirpation Neck dissection Residual tumour after surgery Negative Positive Extent of irradiation field N + PPTS NU Irradiation tumour dose (Gy) ,50 ≥50

No. of patients

5-year DSS

Fig. 3. Disease-specific survival according to (A) extracapsular tumour spread and (B) the extent of the irradiation field (N + PPTS, neck and potential primary tumour sites; NU, neck-unilateral).

%

P-value

tion tumour dose did not influence DSS significantly (Table 2).

6 37 13

100 67 44

0.15

3.2. Causes of death

33 23

73 56

0.21

27 22

79 58

0.41

11 37

100 55

0.01

8 48

75 64

0.43

45 11

71 45

0.09

3.3. Control of neck disease

48 8

71 37

0.04

7 49

50 68

0.10

Neck failure was registered in 10 (18%) patients. In nine of them, it was diagnosed within the first 12 months (median 4 months) after completion of the therapy. Later, at 38 months, neck recurrence was established in one patient only. In all but one patient (pN2B disease with extracapsular spread, irradiated by the neck-unilateral technique), these

N + PPTS, neck and potential primary tumour sites; NU, neck-unilateral.

At the end of the observation time, 23 (41%) patients were alive without evidence of disease, 19 (34%) patients had died from disease and 14 (25%) patients had died from unrelated causes. Cause-related deaths occurred due to neck failure in 10/19 cases. Three patients died due to the primary tumour and six patients died due to distant metastases without evidence of persistent tumour in the head and neck region. All but two cause-related deaths occurred within the first 44 months after diagnosis (median 10 months); the remaining deaths occurred at 87 and 108 months, both due to the primary lesion.

37

P. Strojan, A. Anicˇin / Radiotherapy and Oncology 49 (1998) 33–40 Table 3 Mucosal site failure: clinical and tumour characteristics of patients who subsequently developed head and neck primary tumour Patient

Nodal disease

Primary tumour

Survival

No.

Age (years)/ sex

pN stage

Surgery

RT treatment volume

RT dose (Gy)

Time to diagnosis (months)

Site

TNM stage

Therapy

Response (WHO)

Follow-up (months)

Status

1 2 3 4 5

50/M 46/M 77/M 45/M 50/M

3 3 2A 2B 1

ND ND ND ND ND

NU N+ N+ N+ N+

60 61 43a 61.8 50.4

21 17 86 98 16

Oropharynx Oropharynx Maxillary sinus Nasopharynx Larynx

T1N0M0 T2N0M0 T3N0M0 T4N0M0 T2N0M0

Surgery RT – RT Surgery

CR PR PD PR CR

73 29 87 108 44

NED DOD DOD DOD DOCb

PPTS PPTS PPTS PPTS

ND, neck dissection; NU, neck-unilateral; N + PPTS, neck and potential primary tumour sites; RT, radiotherapy; CR, complete response; PR, partial response; PD, progressive disease; NED, no evidence of disease; DOD, death due to disease; DOC, death due to other causes. a Radiotherapy discontinued on patient’s demand. b New primary tumour, histologically verified adenocarcinoma of the lung.

were in-field failures. They were located at the site of preexistent nodal disease in seven patients and/or outside of it (i.e. somewhere else in the neck) in two patients. In one patient, disease progression was observed just before the course of radiotherapy on the surgically untreated contralateral side of the neck. Two of the neck recurrences were registered in the group of eight patients in whom extirpation of a single neck metastasis was performed only; both recurrent lesions developed in the surgical scar. The occurrence of neck failure correlated significantly only with the extent of the irradiation field (P = 0.03). When using the unknown primary tumour treatment volume, the rate of neck recurrence was 12% (6/48), while when the neck alone was irradiated it was 50% (4/ 8). Comparing both treatment groups, the latter had a significantly larger proportion of patients with a less favourable nodal stage (pN1, six patients versus 0 patients; pN2, 34 patients versus three patients; pN3, eight patients versus five patients; P = 0.01). However, there was no difference in the median cumulative tumour dose between the groups. 3.4. Primary carcinomas A total of five (9%) head and neck primary cancers were subsequently discovered 16–98 months (median 21 months) after the initial diagnosis. The topographic distribution of the primaries as well as the therapy and survival of patients are presented in Table 3. None of the primary tumours were diagnosed below the clavicles. 3.5. Distant metastases Six (11%) patients developed distant metastases in one or several sites (liver, three patients; bone, two patients; lung, three patients; lymph nodes in other than neck regions, one patient) from 2 to 39 months (median 7 months) following diagnosis. The distribution of patients with disseminated disease according to pN stage was N2A in one patient, N2B in three patients and N3 in two patients. The tumour had spread across the node capsule in all six patients. The

incidence of distant metastatic disease did not correlate significantly with any of the clinical or histopathological characteristics tested. 3.6. Treatment morbidity Treatment-related morbidity has been systematically recorded on regular follow-up examinations and for the purpose of this study it was scored from patients’ records. Surgery-related morbidity was limited to painful sensation and/or reduced mobility of the neck and shoulder on the operated side. It was observed to a various degree in 33 patients; all had radical or extended radical neck dissection. In a group of patients irradiated by the large-field technique, mucositis and dermatitis of grade 3 (according to the RTOG normal tissue toxicity scale [20]) were registered in 27 and three patients, respectively. Grade 4 mucositis occurred in four patients. Late adverse effects were mostly limited to persistent xerostomia and subcutaneous/muscular fibrosis, which were registered in 35 and 22 patients, respectively. Two patients developed trismus. Of those patients irradiated to the neck only, one presented with pronounced soft tissue fibrosis. All patients who were without evidence of disease at the last follow-up developed at least one type of the abovementioned late side-effects.

4. Discussion Although it is a relatively uncommon clinical entity, there is a substantial amount of literature reports referring to cervical lymph node metastases from squamous cell carcinoma of an unknown primary tumour. Despite the long time periods they usually cover, the published series are generally small and hardly comparable due to their heterogeneity arising from patient and tumour characteristics, treatment intent and the modalities used. Consequently, data addressing the optimal treatment approach do not allow definitive conclusions, although the results

38

P. Strojan, A. Anicˇin / Radiotherapy and Oncology 49 (1998) 33–40

Table 4 Results of combined modality therapy for cervical lymph node metastases from an occult primary squamous cell carcinoma: review of the literature Reference (year of publication)

Patient entry No. of patients

Bataini et al. [2] (1987) Lefebvre et al. [14] (1990)

1960–1980 1974–1983

Wang et al. [27] (1990)

1953–1988

Maulard et al. [16] (1992) Davidson et al. [7] (1994)

1972–1986 1977–1990

Nguyen et al. [19] (1994)

1978–1991

Present series (1997)

1975–1994

48 (all surgery + RT) 97 (13 RT + surgery, 84 surgery + RT) 41 (11 RT + surgery, 30 surgery + RT) 113 (all surgery + RT) 54b (5 RT + surgery, 49 surgery + RT) 53b (1 RT + surgery, 52 surgery + RT) 56 (all surgery + RT)

Radiotherapy treatment volume

Mucosal site primary (%)a

Neck failure (%)

Distant metastases (%)

DSS/OS at 5 and 10 years (%)

N + PPTS N + PPTS (majority of patients) N + PPTS

NR NR

17 20

25 NR

NR/55, NR/40 NR/35, NR/19

10c

22

NR

6/NR, NR/NR

N + PPTS N + PPTS (majority of patients) N + PPTS

7 9.6

14.1 26

16 20.5

NR/38, NR/NR 60/45, NR/NR

2

13

20

NR/63, NR/59

N (8), N + PPTS (48)

5

18

11

66/52, 52/22

NR, not reported; RT, radiotherapy; N, neck; N + PPTS, neck and potential primary tumour site(s); DSS, disease-specific survival; OS, overall survival. Discovered within 5 years after therapy. Reported results refer to the entire group of patients reported in the paper (Ref. [7], n = 73; Ref. [19], n = 54). c Primary tumours appearing below the clavicles are not excluded. a

b

of combined surgery and radiotherapy seem to be most advantageous (Table 4). The treatment policy in our departments gives preference to surgery and postoperative radiotherapy whenever feasible. Following elimination of a gross tumour burden and micrometastases with primary neck dissection, a total irradiation dose can be reduced in the majority of cases, which results in less pronounced and acceptable treatment-related morbidity, as was the case in our study. Besides, a histopathological examination of a resected specimen provides additional information about the extent and aggressiveness of disease (pN stage, histopathological grade, extracapsular spread, neoplastic emboli in lymphatic vessels and perineural tumour invasion). Thus, the radiation oncologist can objectively consider assessed prognostic factors, so that under- or overtreatment is less likely. Another advantage of combined therapy is improved control of neck disease. This is clearly evident from the report of Davidson et al. [7]. They compared treatment results calculated in a series of 73 patients, of whom 68% had neck dissection and radiotherapy, with those obtained in historical controls treated mostly with surgery alone. The authors reported a significant improvement in the control of the treated neck for those in the combined therapy group, i.e. 74% versus 50%. Similar results were reported in other comparative studies. Bataini et al. [2] recorded a 17% neck failure rate in the combined modality treatment group versus 43% in the group treated with radiotherapy alone and the failure rates of Jesse et al. [12] were 14% in the surgery/ radiotherapy group, 24% in the surgery alone group and 21% in the radiotherapy alone group. In our study, 10/56 (18%) patients failed locally on the neck, which is consistent with the results of other studies dealing with combined therapy (Table 4). In patients treated with curative intent, we preferred to give elective radiation treatment to all possible primary

tumour sites in the head and neck in order to diminish the incidence of subsequently developed primaries. Which parts of the pharyngeal axis will be encompassed in the irradiation field depends on the patient’s life-style and the level of nodal involvement. It is well documented that in those with histology of squamous cell carcinoma the nasopharynx, tonsil and base of the tongue are the most frequent sites of origin [24] and the manifestation of the primary adversely affects the prognosis [6,15,25]. We registered five patients developing a subsequent head and neck primary tumour. All but one were found to have an in-field carcinoma. In 3/5 patients the appearance of a mucosal lesion was detected within the first 5 years after diagnosis of neck metastases, likely representing a failure to detect the primary tumour initially (Table 3). Carcinomas in the maxillary sinus and nasopharynx were diagnosed at 86 and 98 months, respectively. The carcinoma of the maxillary sinus most likely represented a second metachronous primary tumour, whereas the nasopharyngeal primary could be considered as initially occult origin of neck metastases. Late recurrence of the primary in this cancer type is a known phenomenon [26]. However, the incidence of subsequently discovered primaries did not correlate with the extent of the irradiated field. The literature addressing the effectiveness of elective irradiation of a potential primary in the head and neck is controversial [8]. According to the recommendation by Carlson et al. [5], the entire pharynx should be encompassed in the irradiation field. Irradiation to the neck alone should be considered only in cases with a low probability of a primary along the pharyngeal axis (adenocarcinoma histology, submandibular or submental node), or in patients with a poor performance status. When there is a strong suggestion that the nasopharynx is the primary site, the hypopharynx can be omitted from treatment. Analyzing a series of 93 patients, the authors reported a trend towards better survival

P. Strojan, A. Anicˇin / Radiotherapy and Oncology 49 (1998) 33–40

in the group receiving radiation to the entire pharyngeal axis and the neck (47% DSS at 5 years) compared to those receiving radiation to the naso-oropharynx and the neck or to the neck alone (26% and 20% DSS at 5 years, respectively). It should be pointed out that the low survival rate in the naso-oropharynx and neck group reflects a higher percentage of patients with more advanced disease in the neck. In addition, Jesse et al. [12], Coker et al. [6], Bataini et al. [2] and Lefebvre et al. [14] concluded unanimously that large-field prophylactic radiotherapy appears to be highly effective in diminishing the risk of a subsequent appearance of an initially occult primary lesion. Moreover, Harper et al. [10] did not find any significant difference in the incidence of a second metachronous head and neck carcinoma between patients with unknown primary tumour who received radiotherapy to the neck and at least part of the mucosa of the head and neck, and those irradiated for a known primary squamous cell carcinoma arising in the head and neck region. On the other hand, Weir et al. [28] concluded that for patients in whom there is no strong clinical suspicion of a primary tumour, treatment of the involved neck alone provides a comparable survival outcome to more extensive irradiation of all possible primary tumour sites in the head and neck (48% versus 37% OS at 5 years). The number of subsequently discovered head and neck primaries in their series was insignificantly lower in the group receiving largefield irradiation (1/59) than in the group receiving irradiation to the nodes alone (6/85). Results of the studies by Marcial-Vega et al. [15], Mohit-Tabatabai et al. [18] and Glynne-Jones et al. [9] support the conclusions of Weir et al. [28], making the value of elective irradiation uncertain. Even after combining surgery and radiotherapy there has been no substantial improvement in long-term survival when compared with series utilizing surgery or radiotherapy alone [6,7,12,27]. The 5- and 10-year DSS of 66% and 52%, respectively, and OS of 52% and 22%, respectively, calculated in our series leave much to be desired but are favourably comparable with the results reported in the literature (Table 4). The correlation between the DSS of our patients and extracapsular spread was significant and consistent with some other reports [2,16]. It was also significantly influenced by the extent of the irradiation field, i.e. patients irradiated electively with the large-field technique survived better than those irradiated to the involved neck only. This, at least in part, reflects a significant imbalance in the distribution of nodal stage between both treatment groups. Nevertheless, the pN stage of the disease did not significantly affect the survival of our patients, although higher survival rates were associated with lower stages, as reported by Wang et al. [27], Maulard et al. [16], Davidson et al. [7] and Nguyen et al. [19]. An improved local control and diminished incidence of a subsequently developed primary in the head and neck region highlighted the problem of systemic dissemination as a crucial factor that may decisively influence the survival out-

39

come. In the present series, the incidence of distant metastases was 11%. A substantial proportion of patients developing distant metastases, ranging between 12 and 25%, are also reported by other authors (Table 4). All our patients with distant metastases had extracapsular tumour spread with more advanced pN stages. For high risk patients the addition of systemic therapy to initial local treatment seems to be the only way of improving treatment results. In conclusion, acceptable toxicity, the superior control rate of neck disease and the survival results favourably comparable to those reported elsewhere and obtained by a single modality approach, as observed in our study, strongly support the use of combined therapy. However, in patients with a poor clinical and histopathological profile a more aggressive treatment approach exploiting chemotherapy seems to be justified.

Acknowledgements The authors wish to thank Professors Marjan Budihna and Lojze Sˇmid for their valuable comments and discussions. We also thank Mrs Olga Shrestha for her language proofreading.

References [1] Armitage, P. Statistical Methods in Medical Research. Blackwell, Oxford, 1971. [2] Bataini, J.P., Rodriguez, J., Jaulerry, C., Brugere, J. and Ghossein, N.A. Treatment of metastatic neck nodes secondary to an occult epidermoid carcinoma of the head and neck. Laryngoscope 97: 1080–1084, 1987. [3] BMDP Statistical Software. University of California Press, Berkeley, CA, 1990. [4] Cancer Registry of Slovenia. Cancer Incidence in Slovenia 1994. Report No. 36. Institute of Oncology, Ljubljana, 1997. [5] Carlson, L.S., Fletcher, G.H. and Oswald, M.J. Guidelines for radiotherapeutic techniques for cervical metastases from an unknown primary. Int. J. Radiat. Oncol. Biol. Phys. 12: 2101–2110, 1986. [6] Coker, D.D., Casterline, P.F., Chambers, R.G. and Jaques, D.A. Metastases to lymph nodes of the head and neck from an unknown primary site. Am. J. Surg. 134: 517–522, 1977. [7] Davidson, B.J., Spiro, R.H., Patel, S., Patel, K. and Shah, J.P. Cervical metastases of occult origin: the impact of combined modality therapy. Am. J. Surg. 168: 395–399, 1994. [8] Freeman, D., Mendenhall, W.M., Parsons, J.T. and Million, R.R. Unknown primary squamous cell carcinoma of the head and neck: is mucosal irradiation necessary? Int. J. Radiat. Oncol. Biol. Phys. 23: 889–890, 1992. [9] Glynne-Jones, R.G.T., Anand, A.K., Young, T.E. and Berry, R.J. Metastatic carcinoma in the cervical lymph nodes from an occult primary: a conservative approach to the role of radiotherapy. Int. J. Radiat. Oncol. Biol. Phys. 18: 289–294, 1990. [10] Harper, C.S., Mendenhall, W.M., Parson, J.T., Stringer, S.P., Cassisi, N.J. and Million, R.R. Cancer in neck nodes with unknown primary site: role of mucosa radiotherapy. Head Neck 12: 463–469, 1990. [11] International Union Against Cancer (UICC). Head and neck tumours. In: TNM Classification of Malignant Tumours, 5th edn., pp. 17–50. Editors: L.H. Sobin and Ch. Wittekind. Wiley-Liss, New York, 1997.

40

P. Strojan, A. Anicˇin / Radiotherapy and Oncology 49 (1998) 33–40

[12] Jesse, R.H., Perez, C.A. and Fletcher, G.H. Cervical lymph node metastasis: unknown primary cancer. Cancer 31: 854–859, 1973. [13] Kaplan, E.L. and Meier, P. Nonparametric estimation from incomplete observations. J. Am. Stat. Assoc. 53: 457–481, 1958. [14] Lefebvre, J.L., Coche-Dequeant, B., Van, J.T., Buisset, E. and Adenis, A. Cervical lymph nodes from an unknown primary tumour in 190 patients. Am. J. Surg. 160: 443–446, 1990. [15] Marcial-Vega, V.A., Cardenes, H., Perez, C.A., et al. Cervical metastases from unknown primaries: radiotherapeutic management and appearance of subsequent primaries. Int. J. Radiat. Oncol. Biol. Phys. 19: 919–928, 1990. [16] Maulard, C., Housset, M., Brunel, P., et al. Postoperative radiation therapy for cervical lymph node metastases from an occult squamous cell carcinoma. Laryngoscope 102: 884–890, 1992. [17] Million, R.R., Nicholas, J.C. and Mancuso, A.A. The unknown primary. In: Management of Head and Neck Cancer: A Multidisciplinary Approach, 2nd edn., pp. 311–320. Editors: R.R. Million and N.J. Cassisi. J.B. Lippincott, Philadelphia, PA, 1994. [18] Mohit-Tabatabai, M.A., Dasmahapatra, K.S., Rush, B.F., Jr. and Ohanian, M. Management of squamous cell carcinoma of unknown origin in cervical lymph nodes. Am. Surg. 52: 152–154, 1986. [19] Nguyen, C., Shenouda, G., Black, M.J., Vuong, T., Donath, D. and Yassa, M. Metastatic squamous cell carcinoma to cervical lymph nodes from unknown primary mucosal sites. Head Neck 16: 58– 63, 1994. [20] Perez, C.A. and Brady, L.W. Overview. In: Principles and Practice of Radiation Oncology, 2nd edn., pp. 51–53. Editors: C.A. Perez and L.W. Brady. J.B. Lippincott, Philadelphia, PA, 1992.

[21] Peto, R., Pike, M.C., Armitage, P., et al. Design and analysis of randomised clinical trials requiring prolonged observation of each patient. II. Analysis and examples. Br. J. Cancer 35: 1–39, 1977. [22] Robbins, K.T., Medina, J.E., Wolfe, G.T., Levine, P.A., Sessions, R.B. and Pruet, C.W. Standardizing neck dissection terminology: official report of the Academy’s Committee for head and neck surgery and oncology. Arch. Otolaryngol. Head Neck Surg. 117: 601– 605, 1991. [23] Shah, J.P., Strong, E., Spiro, R.H. and Vikram, B. Surgical grand rounds. Neck dissection: current status and future possibilities. Clin. Bull. 11: 25–33, 1981. [24] Shaw, H.J. Metastatic carcinoma in cervical lymph nodes with occult primary tumour – diagnosis and treatment. J. Laryngol. Otol. 84: 249–265, 1970. [25] Talmi, Y.P., Wolf, G.T., Hazuka, M. and Krause, C.J. Unknown primary of the head and neck. J. Laryngol. Otol. 110: 353–356, 1996. [26] Wang, C.C. Re-irradiation of recurrent nasopharyngeal carcinoma – treatment techniques and results. Int. J. Radiat. Oncol. Biol. Phys. 13: 953–956, 1987. [27] Wang, R.C., Goepfert, H., Barber, A.E. and Wolf, P. Unknown primary squamous cell carcinoma metastatic to the neck. Arch. Otolaryngol. Head Neck Surg. 116: 1388–1393, 1990. [28] Weir, L., Keane, T., Cummings, B., et al. Radiation treatment of cervical lymph node metastases from an unknown primary: an analysis of outcome by treatment volume and other prognostic factors. Radiother. Oncol. 35: 206–211, 1995.