Squamous carcinoma of the posterior pharyngeal wall

Squamous Carcinoma of the Posterior Pharyngeal Wall Ronald H. Spiro, MD, Javier Kelly, MD, Alberto Luna Vega, MD, Louis B. Harrison, MD, Elliot W. Strong, MD, New York, New York

We have reviewed a 12-year experience with 295 patients treated for squamous carcinoma of the pharynx in order to focus on 78 patients whose lesions arose in the posterior wall. Surgery was the definitive therapy for the primary tumor in 57 ( 7 3 % ) , including 3 treatment groups. Thirty-two patients bad limited resections that preserved the larynx, involving local excision (7 patients), anterior pharyngotomy (7 patients), lateral pharyngotomy (6 patients), median labiomandibular glossotomy (6 patients), or median mandibulotomy with paralingual extension (6 patients). The second group consisted of 21 patients with more extensive tumors who required a laryngectomy and complex reconstruction, often with postoperative radiotherapy. Finally, there were four patients who developed metachronons second primaries in the pharynx subsequent to a laryngectomy. All required flap reconstruction. Of the 21 patients whose primary treatment was radiotherapy, 5 had lesions that were implanted after access was provided by a mandibulotomy. Cumulative 5-year survival was 32% and ranged from 44% in those with favorable lesions to 15% in those with extensive tumors. Our experience highlights the variety of treatment approaches available in patients with pharyngeal carcinomas confined to the posterior wall. Surgery in this setting carries acceptable morbidity and yields survival rates that compare favorably with those achieved by external radiation therapy alone. Results in patients with extensive lesions still leave much to be desired, despite radical surgery and aggressive radiotherapy. Innovative brachytherapy techniques using surgery for access deserve further investigation.

he posterior pharyngeal wall is a rare site for squamous carcinoma. Delay in diagnosis is common, tuT mors are often sizable, and survival rates are poor. There have been two previous reports from our hospital concerned with the treatment of pharyngeal wall carcinoma, neither of which analyzed the results according to the site From the Head and Neck Service, Department of Surgery, and the Brachytherapy Service, Department of Radiation Oncology, Memorial Sloan-Kettering Cancer Center, New York, New York. Requests for reprints should be addressed to Ronald H. Spiro, MD, 425 East 67th Street, New York, New York 10021. Presented at the 36th Annual Meeting of the Society of Head and Neck Surgeons, Washington, DC, May 19-22, 1990.

420

THE AMERICAN JOURNAL OF SURGERY

of origin [1,2]. There are some significant differences between pyriform fossa, postcricoid, and posterior pharyngeal tumors that deserve attention. It is for this reason that we have limited this study to patients with posterior wall lesions in order to better assess our growing interest in resection with larynx preservation, adjunctive megavoltage radiotherapy, newer reconstructive methods, and innovative brachytherapy techniques. PATIENTS AND METHODS During the years 1977 through 1988, a total of 295 patients received treatment in our hospital for squamous carcinoma of the pharyngeal wall (excluding the nasopharynx). From these, we selected 78 patients whose lesions arose in the posterior wall. There were 48 men and 30 women (1.6:1) who ranged in age from 41 to 76 years (median: 60 years). No prior treatment had been given to the pharyngeal tumor in 68 of these patients, including 4 who had a laryngectomy previously for another primary. The remaining 10 patients had residual or recurrent tumor after radiotherapy. Dysphagia and odynophagia were the most common presenting symptoms (58 patients), followed by the appearance of a neck mass (14 patients) or hoarseness (7 patients). In seven patients, the tumor was an incidental discovery during the evaluation of, or treatment for, another primary. It is worth noting that other primary tumors were documented in at least 29 patients (37%). The larynx (13 patients) and the oral cavity (9 patients) were the sites most often involved. The tumor was completely confined to the posterior wall in 20 patients, involved the lateral walls in 31, and extended to the pyriform fossae, larynx, postcricoid area, or cervical esophagus in 27. Information was adequate for staging in 65 of 68 previously untreated patients: 6 (9%), 24 (37%), 19 (29%), and 16 (25%) had stage 1 through stage 4 tumors, respectively (Figure 1). Localized disease was present in 41 (63%), whereas 24 (37%) had clinical evidence of cervical metastases. The treatment modalities used are summarized in Table I. Surgery was the primary modality in 59 patients (76%), and 19 had definitive radiotherapy (24%). Most often, a multimodality approach was used. In Table II, the choice in previously untreated patients is analyzed according to the stage of the tumor. Limited operations that preserved the larynx and did not include a circumferential pharyngectomy were performed in a total of 32 patients (54% of those treated surgically). Twenty-five had radical operations involving a total laryngectomy or a circumferential pharyngectomy. Table III lists the surgical procedures in detail, excluding two patients who had a neck dissection only. Radiotherapy was the definitive treatment in a total of

VOLUME 160

OCTOBER 1990

CARCINOMA OF THE POSTERIOR PHARYNGEAL WALL

T1

T2

NO

T3

T4

9

2

N1

2

4

4

3

N2

1

3

1

3

N3

0

0

1

2

Figure 1.

Patient distribution by T and N categories according to the AJC classification. Stage groupings are s h o w n by the appropriate blocks.

19 patients. This included 13 patients who received only external beam therapy and 6 who had a combination of brachytherapy and teletherapy. There were four patients in the latter group who had a mandibulotomy approach (either paralingual or translingual) to expose the primary for the implant. Of the total of 29 patients who presented with clinical evidence of cervical node involvement, 20 had a radical neck dissection and 9 received radiotherapy. Nodes were not enlarged initially (NO) in 41 of 65 previously untreated, adequately documented patients. An elective lymphadenectomy was performed in 20 (including 1 modified and 2 selective neck dissections), 4 had beam radiotherapy, and 17 (26%) were observed. In 30 of the 57 patients (53%) who had a resection of their pharyngeal tumor, an immediate reconstruction was performed. The type of repair varied according to the extent of the resection and the preference of the surgeon (Table IV). Reconstruction was required in 15 of 35 patients with T1 or T2 primaries (43%), and 16 of the remaining 22 with larger or recurrent tumors (73%). Survivorship was estimated using the direct and the Kaplan-Meier methods [3]. For the former, patients eligible for 2-year analysis who remained alive and well were compared with the total eligible. Net or determinate 2-year "cure" excluded those who were tumor-free when

TABLE I

Treatment

Surgery only Surgery + postoperative radiotherapy Surgery + chemotherapy + postoperative radiotherapy Radiotherapy only Radiotherapy -I- chemotherapy Radiotherapy + neck dissection Radiotherapy (with surgical access)

Primary Therapy (n = 88)

Previous Therapy (n = 10)

Total (%) (n = 78)

26 22

9 --

35 (45) 22 (28)

2

--

2 (3)

9 5 1

----

9 (12) 5 (6) 1 (1)

3

1

4 (5)

either they died for other reasons or they were lost to follow-up. Survival curves were compared using the logrank test [4]. RESULTS Significant treatment complications were recorded in 39 patients (50%), one of whom died postoperatively. The complication rate was 53% in surgically treated patients and 42% in those receiving definitive radiotherapy. Pharyngocutaneous fistulas occurred in 13 patients, 11 of whom had major reconstructive procedures. Fistulas developed in 5 of 6 patients who had deltopectoral flaps, 3 of 6 who had pectoralis major myocutaneous flaps, 1 of 2 who had a free revascularized flap repair, and 2 of 12 whose continuity was restored by a gastric transposition. Wound infections (11 patients) and pulmonary sepsis (14 patients) were the other complications most frequently encountered. Using the direct method for 68 eligible patients, the absolute, 2-year disease-free survival was 32% (22 patients). Excluding 11 indeterminate patients, the net survival was 39% (22 of 57). Actuarial survival for all patients at 2 years and 5 years was 49% and 32%, respectively (Figure 2). Results were better in those who had lesions that were mostly confined to the posterior wall (58% and 34%), and those whose nodes were clinically

TABLE I I

Treatment SelecUon by Stage*

Limited resection Limited resection + postoperative radiotherapy Radical resection only Radical resection + postoperative radiotherapy Radical resection + chemotherapy + postoperative radiotherapy Radiotherapy only Radiotherapy + chemotherapy Radiotherapy + neck dissection

Stage 1 (n = 6)

Stage 2 (n = 24)

Stage 3 (n = 19)

Stage 4 (n = 16)

3 2 0 0 0 1 0 0

8 5 5 3 0 2 0 1t

2 5 5 1 2 2 1 1

0 3 0 4 0 4 3 2t

9 Sixty-five patients with adequate data for staging. t Surgical access for implant.

THE AMERICAN JOURNAL OF SURGERY

VOLUME 160

OCTOBER 1990

421

SPIRO ET AL

TABLE I I I

TABLE IV

Resection of the Primary Tumor In 57 Patients

Immediate Reconstruction in 31 of 55 Patients*

Patients Procedures

n

%

Local excision Transhyoid or anterior pharyngetomy Median labiomandibular glossotomy Mandibulotomy, paralingual extension Lateral pharyngotomy Circumferential pharyngectomy Pharyngoesophagectomy Pharyngolaryngectomy Pharyngolaryngoesophagectomy

7 7 6 6 6 3* 1* 9 12

12 12 11 11 11 5 2 16 21

Tumor Stage T3 T4 Recurrent

T1

T2

STSG D e l t o p e c t o r a l flap

1 0

4 4

0 2

0 0

0 0

Total 5 6

PMM flap Pull-up ~ Free flap ~;

0 0 0

3 1 2

2 6 0

0 1 0

1 4 0

6 12 2

* Excludes two patients who had only neck dissection. t Gastric tranposition. One radial forearm and one jejunal flap. PMM = pectoralis major myocutaneous; STSG = split thickness skin graft.

* Prior laryngectomy.

'rop Surviving

Proportion Surviving 0.8

0.8

n-30 0.6

0.6

0.4 0.4

49%

0.2

0.2

32%

0 0

i 12

i 24

i 36

i 48

r 60

0 - - - = 0 i___J 72

J 12

36

48

i 6O

Time in Mos

84

Time in Months

24

--

STAGE 1,2

--+-- STAGE 3

--~-- STAGE 4

Figure 2. Cumulative survival for all 78 patients. Forty-nine percent and 32% were alive 2 and 5 years after treatment, respectively.

Figure 3. Survival according to clinical stage in 65 previously untreated patients whose charts contained adequate information. Only the difference between stages 1 and 2 combined, and stage 4 was significant (p <0.01).

negative on admission (61% and 44%). Survival was greater in those with stage 1 and stage 2 tumors, but the difference was significant only in comparison to those with stage 4 disease (Figure 3). Neither age, gender, nor prior treatment status seemed to have an impact. Local recurrence occurred in at least 27 patients (41% of determinate patients) and was directly related to tumor stage (16%, 21%, 37%, and 63% in stages 1 through 4, respectively). Distant metastasis, almost always pulmonary, was documented in six patients only. Neck failure was noted in a total of 14 patients (21% of determinate patients). The latter included 7 of 20 patients who had a therapeutic lymphadenectomy, 2 of 9 who had irradiation of clinically positive nodes, 3 of 20 who had an elective lymphadenectomy, and 2 of 20 patients who had a clinically negative neck and who received no treatment initially.

3% to 25% using varying methods and time intervals. Our results seem better, but it should be noted that 63% of our patients had localized disease initially, compared with 44% to 54% in the other studies. When the survival of our patients was analyzed by stage, the differences were not as significant as would be expected. This may be related to the current AJC staging system, which classifies oropharyngeal tumors according to size and hypopharyngeal primaries by extent [10]. Staging of posterior pharyngeal wall lesions can be a problem as these tumors not infrequently involve both regions of the pharynx. It is worth noting that the proportion of women with pharynx cancer treated in our hospital has increased rather dramatically from 19% to 38% during the past four decades. Presumably, changing patterns in the use of alcohol and tobacco explain the unusually high incidence of women included in this study. There is general agreement that the limited effectiveness of radiotherapy alone relates to the technical problem posed by the close proximity of the radiation-sensitive spinal cord. Even when this problem is surmounted and adequate doses are delivered, the fact remains that tu-

COMMENTS Studies that specifically focus on posterior pharyngeal wall cancer have been few, but the low incidence of these tumors and the poor results of treatment are well documented [5-9]. Reported survival rates have ranged from 422

THE AMERICAN JOURNAL OF SURGERY VOLUME 160 OCTOBER1990

CARCINOMAOF THE POSTERIOR PHARYNGEALWALL

mors arising in the oropharynx or hypopharynx do not respond like their counterparts in the nasopharynx. The reason for this difference remains unknown, but may relate in part to the lower incidence of undifferentiated tumors in the posterior pharyngeal wall. Treatment comparisons can be misleading in any retrospective study. The choice of surgery as the definitive treatment in 76% of our patients clearly reflects a longstanding preference for this modality in our hospital. Aside from a well-established commitment to postoperative radiotherapy in recent years, what seems noteworthy is that larynx preservation was possible in 54% of our patients who had resections. Whether the approach was peroral, transhyoid, transmandibular, or by way of a lateral pharyngotomy, adequate oral alimentation was usually achieved. Increasing confidence in larynx-sparing surgery has led us to more aggressive procedures for access, such as a median mandibulotomy with paralingual extension, in conjunction with a lateral pharyngotomy. As would be anticipated, survival is not compromised when a normal, uninvolved larynx is preserved. Although we agree with McNeill [8] that preservation of the internal laryngeal nerve, laryngeal suspension, and cricomyotomy may help to minimize aspiration, none of these measures seems to be essential. When a primary repair is not possible, several excellent reconstructive options are available. A pectoralis major myocutaneous flap or a free, revascularized radial forearm flap can be used to repair defects of any size when some of the pharynx remains. Following circumferential pharyngectomy, continuity can immediately be restored either with the transposed stomach or free, revascularized jejunum. The choice of reconstruction obviously depends on experience and preference, and this study confirms that significant morbidity can be anticipated regardless of the method used. It seems clear that better control of the primary tumor must be achieved if results are to improve. This assumption must be tempered by the reality that 37% of our patients had additional primary tumors. We suspect that the low incidence of distant metastasis observed in our patients directly relates to the low survival rate associated with poor control of the primary tumor and a high incidence of second cancers. Our experience also mirrors the comment of Marks et al [I 1] that the incidence of cervical metastases from posterior pharyngeal wall cancer is lower than that observed in patients with pyriform fossa,

base of tongue, or tonsil tumors. Most of the treatment failures in the neck occurred in patients who initially had palpable metastases. The impact of elective neck therapy seemed minimal. The role of radiotherapy and surgery in the treatment of squamous carcinoma of the posterior pharyngeal wall remains in flux. It seems clear that results after aggressive use of both modalities still leave much to be desired, particularly in those patients who have large primaries and bulky nodal disease. Prospective trials using adjunctive chemotherapy would seem appropriate in this setting. Excellent local control has been reported in a small, select group of patients with pharyngeal wall cancers who received peroral implants and adjunctive, megavoltage irradiation [12]. We have expanded this concept to include patients whose tumors were not accessible for implantation through the open mouth. Our limited experience with brachytherapy facilitated by surgical exposure suggests that this novel approach deserves further investigation. REFERENCES 1. Cunningham MP, Catlin D. Cancer of the pharyngeal wall. Cancer 1967; 20: 1859-66. 2. Shah JP, Shaha AR, Spiro RH, Strong EW. Carcinoma of the hypopharynx. Am J Surg 1972; 132: 439-43. 3. Kaplan E, Meier P. Nonparametric estimation for incomplete observation. J Am Stat Assoc 1958; 53: 457-81. 4. Mantel N. Evolutionof survival data and two new rank order statistics arising on its consideration.Cancer Chemother Rep 1966; 50: 103-70. 5. Wilkins SA. Carcinoma of the posterior pharyngeal wall. Am J Surg 1971; 122: 477-81. 6. Wang CC. Radiotherapeutic management of carcinoma of the posterior pharyngeal wall. Cancer 1971; 27: 894-6. 7. Pene F, Avedian V, Eshwege F, Barrett A, et al. A retrospective study of 131 cases of carcinoma of the posterior pharyngeal wall. Cancer 1978; 42: 2490-3. 8. McNeill R. Surgical management of carcinoma of the posterior pharyngeal wall. Head Neck Surg 1981; 3: 389-94. 9. Teichgraeber .IF, McConnel FM. Treatment of posterior pharyngeal wall carcinoma. Otolaryngol Head Neck Surg 1986; 94: 287-90. 10. American Joint Committee on Cancer. Manual for staging of cancer. 3rd ed. Philadelphia: JB Lippincott, 1988: 34. 11. Marks ,IE, Smith PG, Sessions DG. Pharyngeal wall cancer, a reappraisal after comparisonof treatment methods.Arch Otolaryngol 1985; 111: 79-85. 12. Son YH, Kacinski BM. Therapeutic concepts of brachytherapy/megavoltagein sequence for pharyngeal wall cancers. Cancer 1987; 59: 1268-73.

THE AMERICAN JOURNAL OF SURGERY VOLUME160 OCTOBER1990 423