Health care delivery for head-and-neck cancer patients in Alberta: a practice guideline

BACKGROUND

The treatment of head-and-neck cancer is complex and requires the involvement of various health care professionals with a wide range of expertise. We describe the process of developing a practice guideline with recommendations about the organization and delivery of health care services for head-and-neck cancer patients in Alberta.

METHODS

Outcomes of interest included composition of the health care team, qualification requirements for team members, cancer centre and team member volumes, infrastructure needs, and wait times. A search for existing practice guidelines and a systematic review of the literature addressing the organization and delivery of health care services for head-and-neck cancer patients were conducted. The search included the Standards and Guidelines Evidence (sage) directory of cancer guidelines and PubMed.

RESULTS

One practice guideline was identified for adaptation. Three additional practice guidelines provided supplementary evidence to inform guideline recommendations. Members of the Alberta Provincial Head and Neck Tumour Team (consisting of various health professionals from across the province) provided expert feedback on the adapted recommendations through an online and in-person review process. Selected experts in head-and-neck cancer from outside the province participated in an external online review.

SUMMARY

The recommendations outlined in this practice guideline are based on existing guidelines that have been modified to fit the Alberta context. Although specific to Alberta, the recommendations lend credence to similar published guidelines and could be considered for use by groups lacking the resources of appointed guideline panels. The recommendations are meant to be a guide rather than a fixed protocol. The implementation of this practice guideline will depend on many factors, including but not limited to availability of trained personnel, adequate funding of infrastructure, and collaboration with other associations of health care professionals in the province.

The relationship between lingual and hypoglossal nerve function and quality of life in head and neck cancer

Sensorimotor impairment of the tongue has the potential to affect speech and swallowing. The purpose of this study was to critically examine the effects of nerve preservation and reinnervation after reconstruction of the base of tongue on patient-perceived outcomes of quality of life (QoL) related to speech and swallowing through completion of the EORTC QLQ-H&N35 standardised questionnaire. Thirty participants with a diagnosis of base of tongue cancer underwent primary resection and reconstruction with a radial forearm free flap, which may or may not have included nerve repair to the lingual nerve, hypoglossal nerve or both. Eight QoL domains sensitive to changes in motor and sensory nerve function were included in the analysis. Transected lingual and hypoglossal nerves were associated with difficulty in swallowing, social eating, dry mouth and social contact. There were fewer problems reported when these nerves were either repaired or left intact. There were no significant differences between patient nerve status and QoL outcomes for speech, sticky saliva and use of feeding tubes. This study was the first to examine the impact of sensory or motor nerve transection and reconstruction on health-related QoL outcomes.

The impact of a skilled nursing facility on the cost of surgical treatment of major head and neck tumors

BACKGROUND

The finite resources available for health care and the proliferation of managed care in the United States have forced the head and neck surgeon to critically evaluate the cost of tumor treatment.

OBJECTIVE

To determine whether the cost of treating patients with head and neck tumors would be reduced if the patients were to spend a portion of what would otherwise be acute care hospital days in a hospital-based skilled nursing facility (HB/SNF).

DESIGN

Retrospective cost-benefit analysis.

SETTING

Tertiary referral center.

PATIENTS

Twenty-four consecutive hospital admissions for definitive surgical treatment of head and neck tumors were retrospectively reviewed. The postoperative day on which the patient theoretically could have been transferred to the HB/SNF was determined. The charges and cost of each patient's actual hospital stay were compared with the theoretical counterparts had the patient been transferred to the HB/SNF on the determined day.

MAIN OUTCOME MEASURE

Cost savings.

RESULTS

The total hospital stay for the 24 patients was 524 days. One hundred eighty-two of those days could have been spent in the HB/SNF. The total charge and cost savings with the use of an HB/SNF were $201,045 and $84,238, respectively (15% of the total charge and cost). This represents an average charge and cost savings of $8377 and $3510, respectively, per patient. The difference was statistically significant (P<.005).

CONCLUSION

An HB/SNF could reduce the cost of head and neck tumor treatment without compromising patient care.

Cytologic evaluation of surgical irrigation fluid following resection of head and neck squamous cell carcinoma

INTRODUCTION

Historically, squamous cells exfoliated from head and neck carcinoma resection have been implicated in locoregional recurrence, but there have been few studies demonstrating the presence of these cells. This study was designed to evaluate the presence of exfoliated malignant cells in surgical irrigation fluid collected during head and neck cancer resection.

METHODS

Thirty patients undergoing surgery for biopsy-proven squamous cell carcinoma had their surgical sites irrigated with 1,000 cc of normal saline. Surgical gloves and instruments were also washed. These samples were prepared and stained using standard squamous cell cytologic stains. All cases were reviewed by one cytopathologist.

RESULTS

Eighteen patients (60%) had positive or suspicious cytology detected in at least one of the surgical samples. In patients with T0 and T1 tumours, all surgical samples were negative. Positive or suspicious cytology was detected in the primary site and glove and instrument irrigation in 40% of patients with T2 tumours, 42% of patients with T3 tumours, and 50% of patients with T4 tumours. This was statistically significant (p < .05). Positive or suspicious cytology was detected in the neck and glove and instrument irrigation in 29% of patients with an N0 neck, 31% of patients with an N1 neck, 39% of patients with an N2 neck, and 100% of patients with an N3 neck. Five of seven patients (71%) with previous radiation therapy had positive or suspicious cytology in at least one of the surgical samples.

CONCLUSIONS

Higher tumour and nodal staging and a previous history of radiation therapy are associated with an increased incidence of positive or suspicious cytology in surgical irrigation fluid. These findings have implications for surgical protocols.

Submandibular gland transfer: a new method of preventing radiation-induced xerostomia

OBJECTIVE

Radiation-induced xerostomia is a significant morbidity of radiation therapy in the management of patients with head and neck cancers. We have recently reported a method of transfer of one submandibular gland to the submental space in a small pilot series of eligible surgical patients. The submental space was shielded during postoperative radiation therapy. The transferred gland continued to function after the completion of radiation therapy and none of the patients developed xerostomia. The purpose of this article is to present the technique of submandibular gland transfer in detail and to evaluate the postoperative survival and function of the transferred submandibular glands.

DESIGN

Prospective clinical trial.

METHODS

The submandibular gland was transferred on eligible patients as part of their surgical intervention. The patients were followed clinically, with salivary flow and radioisotope studies.

RESULTS

We performed the surgical transfer of the submandibular salivary gland in 24 of 25 patients placed on the protocol. All the glands survived transfer and functioned well postoperatively as demonstrated on the salivary flow and the radioisotope studies. The surgical transfer was relatively simple and added 45 minutes to the surgical procedure. There were no complications attributed to the submandibular gland transfer.

CONCLUSIONS

We have successfully demonstrated that the submandibular gland can be surgically transferred to the submental space with its function preserved. The gland seems to continue functioning even after radiation therapy with the appropriate shielding. This surgical transfer procedure has the potential to change the way we currently manage patients with head and neck cancer.

Subglottic stenosis secondary to extramedullary hematopoiesis in a patient with postpolycythemia myeloid metaplasia

BACKGROUND

Extramedullary hematopoiesis (EMH) is known to occur in myeloproliferative disorders and hemoglobinopathies and is usually seen in the spleen and liver.

METHODS

We report the first case of EMH causing subglottic stenosis in a woman with postpolycythemia myeloid metaplasia (PPMM). A tracheotomy was performed to maintain the airway and local radiotherapy was given.

RESULTS

Two months after the radiotherapy was completed laryngoscopy showed an unobstructed airway with no evidence of disease, and the patient was successfully decanulated. Magnetic resonance imaging 8 months after radiotherapy confirmed the absence of local disease.

CONCLUSION

Consideration should be given to EMH as a possible cause of airway obstruction in the differential diagnosis of a patient with a history of PPMM.

Submandibular salivary gland transfer prevents radiation-induced xerostomia

BACKGROUND

Xerostomia is a significant morbidity of radiation therapy in the management of head and neck cancers. We hypothesized that the surgical transfer of one submandibular salivary gland to submental space, outside the proposed radiation field, prior to starting radiation treatment, would prevent xerostomia.

METHODS

We are conducting a prospective clinical trial where the submandibular gland is transferred as part of the surgical intervention. The patients are followed clinically, with salivary flow studies and University of Washington quality of life questionnaire.

RESULTS

We report early results of 16 patients who have undergone this procedure. Seven patients have finished and 2 patients are currently undergoing radiation treatment. In 2 patients, no postoperative radiation treatment was indicated. Two patients are waiting to start radiation treatment and 2 patients refused treatment after surgery. The surgical transfer was abandoned in 1 patient. All of the transferred salivary glands were positioned outside the proposed radiation fields and were functional. The patients did not complain of any xerostomia and developed only minimal oral mucositis. There were no surgical complications.

CONCLUSIONS

Surgical transfer of a submandibular salivary gland to the submental space (outside the radiation field) preserves its function and prevents the development of radiation-induced xerostomia.

Effect of diameter of microvascular interposition vein grafts on vessel patency and free flap survival in the rat model

Interposition vein grafting is an important technique in microvascular free tissue transfer. Studies in rats have demonstrated that the patency rate of vessels is not affected by interposition grafting when the vein grafts and receipt vessels are of similar diameter. Size discrepancy between vein grafts and recipient vessels is frequently encountered in clinical practice and may potentially be an important factor in anastomotic patency. This study was, therefore, designed to assess the effect of vein graft diameter on the patency of arterial repair and survival of a groin free flap in the rat model. Forty-nine Sprague-Dawley rats were used. The inferior epigastric and femoral veins were used to reconstruct the femoral artery in situ (12 rats, 24 anastomoses) and in groin free flaps (30 rats). The vessel patency with inferior epigastric (1:1 size match) and femoral (2:1 size match) veins was 100% in the non free flap model. In the free flap model, flap survival was 30% in the femoral (2:1 size match) vein graft group. This was significantly less than both the free flap epigastric vein graft group (90% survival) and primary anastomoses group (100% survival). The results of this study suggest that size-matched interposition vein grafts can provide a high degree of reliability, but with size mismatch vein grafts are prone to thrombus formation and subsequent free flap failure.

Role of screening chest computed tomography in patients with advanced head and neck cancer

OBJECTIVE

The presence of pulmonary metastases significantly alters the treatment of patients with head and neck cancers. Currently, a chest radiograph (CXR) is used as a screening examination, although a chest CT (CCT) can detect smaller lesions. The aim of this study was to evaluate the benefit of CCT as a screening tool in patients with newly diagnosed advanced head and neck cancers.

METHOD

New patients with stage III and IV head and neck squamous cell carcinomas were enrolled in this prospective study from August 1994 to December 1995. Twenty-five patients underwent CXR ($71) and CCT ($597) within 2 weeks of diagnosis of the index cancer.

RESULTS

In 20 patients neither the CXR nor the CCT showed any evidence of pulmonary malignancy. Two patients had normal CXRs but possible metastases on CCT. Both the pulmonary lesions resolved on follow-up evaluation. Two patients had suspicious lesions on CXR, 1 of whom had a normal CCT. The second patient underwent CT-guided biopsy which was negative for malignancy. Both the CXR and CCT of the final patient, who had a bronchogenic carcinoma, were suspicious.

CONCLUSION

In 2 patients CCT detected suspicious lesions missed on CXR, although neither revealed malignancy. Three patients with suspicious CXRs would have had CCTs anyway. Thus 22 of 25 CCTs done at the additional cost of $13,314 did not add to the sensitivity of the screening for pulmonary metastasis or second lung primary.

An integrated theory of the no-reflow phenomenon and the beneficial effect of vascular washout on no-reflow

OBJECTIVES

No-reflow is failure of perfusion in free tissue transfer despite adequate arterial inflow. The objectives of this study were to construct a theory of interactive mechanisms of the no-reflow phenomenon and to determine whether preischemic vascular washout could increase flap ischemia tolerance.

STUDY DESIGN

The evidence for the role of various mechanisms in the development of no-reflow is reviewed, and an integrated network proposed. A rat-groin free flap model is used to test preischemic vascular washout with normal saline, heparinized normal saline, lactated Ringer's solution, Tis-U-Sol, and Viaspan.

METHODS

The mean ischemia tolerance of this flap without any therapeutic intervention was first determined, using 22 animals. An additional 50 animals were used to compare with the control group the ischemia tolerance of flaps washed out with the above fluids before their ischemic period.

RESULTS

The critical ischemia time 50 (time after which half of the flaps are expected to survive and half, die) of the untreated flap is 23.4 hours in this model (P<.05). Flaps washed out with normal saline or lactated Ringer's solution have significantly worse ischemia tolerance (P<.0001). Flaps washed out with Tis-U-Sol or Viaspan behave similarly to the control group (P>.57). Flaps receiving preischemic washout with heparinized normal saline (4,000 units/L) had a significantly better outcome than the control group (P<.027).

CONCLUSIONS

Preischemic washout with normal saline, lactated Ringer's solution, or heparinized Tis-U-Sol is detrimental for flap survival after ischemia, Tis-U-Sol- and Viaspan-treated flaps do have ischemia tolerance similar to the control group, and flaps washed out with heparinized normal saline have a survival advantage in this model.

Combined Epstein-Barr virus and human papillomavirus infection in nasopharyngeal carcinoma

Epstein-Barr virus (EBV) has been shown to be a likely etiologic agent in nasopharyngeal carcinogenesis. Human papillomaviruses (HPVs) have previously been identified in numerous upper aerodigestive tract carcinomas. This pilot study was undertaken to investigate the prevalence of combined EBV and HPV infection in 17 patients with nasopharyngeal carcinoma (NPCA) using polymerase chain reaction (PCR). The primary goal was to determine if the presence of HPV could be correlated with molecular, histologic, or clinical parameters. There were seven patients with undifferentiated NPCA (World Health Organization [WHO] type III) and 10 patients with squamous cell carcinoma (WHO type I). All 17 patients had stage IV disease at presentation. EBV was identified in 15 patients (88.2%), and HPV subtypes were identified in samples from nine patients (52.9%). All HPV-positive cases were also EBV positive. Western blot analysis of six samples showed a high level of expression of c-myc and cdc2 kinase and a low level of p53 protein in NPCAs that contained both HPV and EBV (n = 3). Increased expression of c-myc and cdc2 kinase was seen in the cases that contained EBV only, but to a lesser extent (n = 2). These findings indicate an effect of the virus on cellular proliferation and differentiation. Similarly, an elevated level of Rb protein was found only in the HPV-containing NPCAs. Moderate differentiation (keratinization) occurred in four of eight HPV-negative and none of the nine HPV-positive NPCAs. (All HPV-positive cases were poorly differentiated or undifferentiated.) This difference is statistically significant for this sample size (P < 0.03). There was a trend for the group that was HPV positive to have WHO III histology and for the HPV-negative group to have WHO I. The presence of HPV could not be correlated with any clinical parameters in this small group of patients with advanced disease; however, these data suggest that coexistence of EBV and HPV infection may be a factor in the pathogenesis of NPCA and may have an effect on regulation of cellular proliferation and differentiation.

Teaching paradigm for decision making in facial skin defect reconstructions

OBJECTIVE

To present a decision paradigm for facial defect reconstruction, and test the ability of this paradigm to improve resident performance.

DESIGN

A decision paradigm for reconstruction of facial skin defects is proposed and explained, with patient examples. The paradigm's usefulness is then tested with residents.

SETTING

Otolaryngology residency training program at a tertiary hospital.

STUDY PARTICIPANTS

Otolaryngology residents.

INTERVENTIONS

Twelve residents took a pretest wherein they were presented with drawings of skin defects and asked to choose the "best" (most aesthetically pleasing) type of reconstruction from a closed set. This paradigm was presented to these residents, and their posttest consisted of choosing again with the same defects and closed set of choices.

MAIN OUTCOME MEASURES

Cosmetic outcomes of reconstructive decisions on the pretest and posttest were rated on a scale of 0 to 5 (with 0 indicating poor; 5, excellent).

RESULTS

There was a significant improvement in reconstructive choices between the pretest and posttest (P<.001, Student t test).

CONCLUSION

This paradigm can be easily modified to accommodate different surgical approaches preferred by individual surgeons and is thus useful in almost any reconstructive teaching situation.

The Clavipectoral Osteomyocutaneous Free Flap

Microvascular free tissue transfer has revolutionized head and neck reconstruction and currently is considered the most successful and reliable method of primary oromandibular reconstruction. This study was designed to assess the feasibility of full thickness free vascularized transfer of the clavicle based on the clavicular branch of the thoracoacromial artery and the soft tissue component associated with the thoracoacromial axis. Forty dissections of the pectoral region were performed on 26 cadavers. The anatomic relations of the region and the thoracoacromial arterial and venous systems were documented in detail. Selective ink injections of the thoracoacromial arterial branches were also performed on fresh cadavers. The clavicle was supplied mainly by the clavicular artery (medial three quarters), with minor contribution from the deltoid artery (lateral quarter). An average of 16.1 cm (range of 12 to 20 cm) was obtained with total clavicular harvest and the clavicle had sufficient width and height to support dental implants. Two soft tissue donor sites were associated with the thoracoacromial artery: the sternocostal head of the pectoralis major muscle, with the overlying skin supplied by the pectoral artery, and the clavicular head of the pectoralis major muscle, with the overlying skin supplied by the deltoid and clavicular arteries. Sensory innervation of the upper chest was supplied through the supraclavicular nerves, whereas the lateral pectoral nerve supplied motor innervation to both heads of the pectoralis major muscle. The anatomy of the clavipectoral donor site and the first case of full thickness free clavicular transfer for mandibular reconstruction in the English literature are presented. The donor site is an excellent source of well vascularized, thin, pliable, hairless, potentially innervated (motor and sensory) soft tissue, along with up to 20 cm of clavicular bone. The surgical anatomy is familiar to the head and neck surgeon. The harvesting does not require repositioning of the patient and is amenable to a two-team, simultaneous approach. The functional and cosmetic donor site morbidity is minimal even with clavicular harvest. The major disadvantage of this flap is the relatively short pedicle. The authors conclude that the thoracoacromial system provides a free flap with osseous and soft tissue components that are well suited for oromandibular reconstruction.

The clavipectoral osteomyocutaneous free flap

Microvascular free tissue transfer has revolutionized head and neck reconstruction and currently is considered the most successful and reliable method of primary oromandibular reconstruction. This study was designed to assess the feasibility of full thickness free vascularized transfer of the clavicle based on the clavicular branch of the thoracoacromial artery and the soft tissue component associated with the thoracoacromial axis. Forty dissections of the pectoral region were performed on 26 cadavers. The anatomic relations of the region and the thoracoacromial arterial and venous systems were documented in detail. Selective ink injections of the thoracoacromial arterial branches were also performed on fresh cadavers. The clavicle was supplied mainly by the clavicular artery (medial three quarters), with minor contribution from the deltoid artery (lateral quarter). An average of 16.1 cm (range of 12 to 20 cm) was obtained with total clavicular harvest and the clavicle had sufficient width and height to support dental implants. Two soft tissue donor sites were associated with the thoracoacromial artery: the sternocostal head of the pectoralis major muscle, with the overlying skin supplied by the pectoral artery, and the clavicular head of the pectoralis major muscle, with the overlying skin supplied by the deltoid and clavicular arteries. Sensory innervation of the upper chest was supplied through the supraclavicular nerves, whereas the lateral pectoral nerve supplied motor innervation to both heads of the pectoralis major muscle. The anatomy of the clavipectoral donor site and the first case of full thickness free clavicular transfer for mandibular reconstruction in the English literature are presented. The donor site is an excellent source of well vascularized, thin, pliable, hairless, potentially innervated (motor and sensory) soft tissue, along with up to 20 cm of clavicular bone. The surgical anatomy is familiar to the head and neck surgeon. The harvesting does not require repositioning of the patient and is amenable to a two-team, simultaneous approach. The functional and cosmetic donor site morbidity is minimal even with clavicular harvest. The major disadvantage of this flap is the relatively short pedicle. The authors conclude that the thoracoacromial system provides a free flap with osseous and soft tissue components that are well suited for oromandibular reconstruction.

Communication after laryngectomy. An assessment of patient satisfaction

OBJECTIVES

To determine the satisfaction of patients with their current method of alaryngeal communication. To focus primarily on the patients' perception of their own speech.

DESIGN

A retrospective review of patients who underwent total laryngectomy for malignancy identified 4 groups of patients. A survey using a mailed questionnaire was used to compare groups.

SETTING

Tertiary care university hospital, University of Texas Medical Branch at Galveston.

PATIENTS

Forty-seven patients underwent total laryngectomy for cancer and survived. Thirty-one of the 47 patients responded to the survey. Patients were divided into 4 groups by their current method of communication: (1) tablet writers; (2) esophageal speech; (3) electrolarynx; and (4) tracheoesophageal speech.

MAIN OUTCOME MEASURES

Satisfaction with communication, satisfaction with speech quality, ability to communicate over telephone, limitation of interaction with others, and satisfaction with quality of life.

RESULTS

Patients in group 4 were significantly more satisfied with their speech (P < .001), perceived their speech to be of better quality (P < .001), had improved ability to communicate over the telephone (P < .001), and had less limitation of their interactions with others (P < .004). Patients in group 4 also rated their overall quality of life higher (P = .23).

CONCLUSION

Although many studies in the past have demonstrated the objective superiority of tracheoesophageal speech compared with other methods of alaryngeal communication, most studies have focused on the intelligibility of speech judged by listeners. This study demonstrates that patients who use tracheoesophageal speech rate their own speech significantly higher than patients who use other methods and most likely have an overall superior quality of life.

Extended use of topical nasal decongestants

Use of sympathomimetic topical nasal decongestants to treat nasal obstruction is usually restricted to 3 to 5 days to avoid potential rebound swelling (rhinitis medicamentosa). In this study, 10 healthy volunteers used oxymetazoline (long-acting topical nasal decongestant) nightly for 4 weeks. Subjects who used antihistamines, oral or topical decongestants, or systemic steroids or who had active sinusitis were excluded from the study. Weekly history, physical examination, and anterior rhinomanometry revealed no adverse effects. Eight (80%) subjects developed nightly nasal obstruction a few hours before the evening dose; the obstruction resolved within 48 hours if no more decongestant was used. All subjects remained responsive to oxymetazoline 4 weeks and 8 weeks after the study began. This finding suggests that long-acting decongestants may be safely used for longer than the recommended 3 to 5 days without adverse effects if used once nightly.

Biomechanics of the helical rim advancement flap

OBJECTIVES

To determine how helical rim closure with advancement flaps affects ear length and cupping, to compare the tension of closure with wedge excision and helical rim advancement and the effects of 3 tension-decreasing surgical techniques, and to review clinical experience with this flap.

DESIGN

The laboratory study was performed on 6 fresh cadaver ears, by means of sequential excision of tissue, and closure tension was measured with a strain gauge. Results obtained in 10 patients were reviewed.

SETTING

University referral hospital.

PATIENTS

Ten patients with helical rim defects treated with helical rim advancement flaps.

INTERVENTION

In cadaver ears, a helical rim defect of 5 mm was enlarged sequentially to 10 mm, 15 mm, and finally 20 mm. In the patients, defects of the helical rim caused by trauma or tumor were closed by this helical rim advancement flap method.

MAIN OUTCOME MEASURES

For the laboratory study, the outcome measures were tension of closure of the defect, ear length, and ear cupping. For the review of cases, outcome was determination of perioperative complications and the patient's and surgeon's judgment of cosmetic appearance.

RESULTS

Closure of a helical rim defect with advancement flaps caused minor shortening and moderate cupping of the ear. The tension of closure was decreased by extending the inferior incision into the earlobe, creating a Burow triangle, and shaving cartilage from the scapha. Both the Burow triangle and the scaphal shave caused mild increases in ear cupping.

CONCLUSION

Helical rim advancement flaps provide satisfactory closure of helical rim defects up to at least 20 mm (longer in some ears) with excellent preservation of normal anatomic landmarks and a near-normal appearance of the reconstructed ear.

Gastroesophageal reflux prophylaxis decreases the incidence of pharyngocutaneous fistula after total laryngectomy

Pharyngocutaneous fistula is a serious complication after laryngectomy. Gastric fluid is known to cause severe laryngopharyngeal injury and poor mucosal healing. This study was designed to evaluate the effects of a gastroesophageal reflux prophylaxis regimen on the incidence of fistula after laryngectomy. Twenty-one consecutive patients were entered in the study. Patients with positive resection margins were excluded from the overall analysis. All patients had a Connell's two-layer closure of the pharynx with absorbable suture, suction drainage of the neck, postoperative tube feeding, and prophylactic antibiotics. All patients were started on an antireflux regimen postoperatively composed of continuous tube feeds, intravenous ranitidine, and intravenous metoclopramide hydrochloride. Patients were followed postoperatively with Gastrografin swallows and clinically for 8 weeks. The control group consisted of retrospectively studied patients managed identically except for the antireflux prophylaxis. The two groups were well matched for factors reported to influence the rate of pharyngocutaneous fistula formation. The control group had six fistulae (26%) and an average of 16.5 days of hospital stay. The study group had no fistulae and an average of 11.5 days of hospital stay (P = .02). This study suggests that gastroesophageal reflux may predispose to fistula formation after laryngectomy and that mechanical and pharmacological prophylaxsis decreases postoperative morbidity and length of hospital stay.

Infantile subglottic hemangioma

Infantile subglottic hemangiomas are rare congenital lesions of the larynx that are difficult to manage. Our experience with 10 cases was reviewed. All patients presented with stridor at birth, or shortly after. The minimum follow-up was 6 months, with an average of 44 months. The patients were classified and treated according to the size of the hemangioma. The treatment included expectant management, systemic steroids, CO2 laser excision, and tracheostomies. Overall, eight patients were cured, one developed subglottic stenosis, and one died due to unrelated causes. We recommend that (1) patients with small lesions (< 25% of lumen) be treated expectantly, (2) patients with moderate lesions (25-60% of lumen) be treated with systemic steroids initially, and CO2 laser excision if steroids fail, and (3) patients with large lesions (> 60% of lumen) be treated with tracheostomies initially, and staged CO2 laser excision.

Pulmonary atelectasis after reconstruction with pectoralis major flaps

Forty-four patients were reviewed to determine the incidence of atelectasis following pectoralis major myocutaneous flap reconstruction of head and neck defects. Patients underwent tumor resection with subsequent pectoralis major myocutaneous flap reconstruction (flap group, n = 24) or another major head and neck procedure (control group, n = 20). Chest roentgenograms taken on the first postoperative day were scored for atelectasis by preestablished criteria. Sixty-five percent of control and 70% of flap patients demonstrated postoperative atelectasis roentgenographically. The flap patients with skin paddles larger than 40 cm2 had a 60% incidence of major atelectasis compared with 5% in control patients. The skin island area was strongly correlated with the atelectasis score in the flap group. These results suggest that atelectasis is common following pectoralis major myocutaneous flap reconstruction of head and neck defects. As well, decreased chest wall compliance after primary closure of large donor defects may contribute to the atelectasis observed.