- Email: [email protected]
Primary versus secondary tracheoesophageal puncture in salvage total laryngectomy following chemoradiation
Otolaryngology–Head and Neck Surgery (2009) 140, 386-390
ORIGINAL RESEARCH—HEAD AND NECK SURGERY
Primary versus secondary tracheoesophageal puncture in salvage total laryngectomy following chemoradiation Kevin S. Emerick, MD, Luke Tomycz, MD, Carol R. Bradford, MD, Teresa H. Lyden, MA, CCC-SLP, Douglas B. Chepeha, MD, Gregory T. Wolf, MD, and Theodoros N. Teknos, MD, Ann Arbor, MI; Nashville, TN; and Columbus, OH OBJECTIVE: To compare the rate of postoperative wound-healing complications and voice fluency in primary vs secondary tracheoesophageal puncture (TEP) following chemoradiation. METHODS: Between 1998 and 2005, 30 patients underwent laryngectomy after chemoradiation therapy. Twenty patients underwent primary TEP and 10 patients underwent secondary TEP. Comorbidities, postoperative complications, speech fluency, and time to speech fluency were evaluated in each patient. RESULTS: Pharyngocutaneous fistula (PCF) occurred in 10 of 20 (50%) patients who underwent primary TEP and in 0 of 10 (0%) patients in the secondary TEP group (P ⬍ 0.05). Overall, 25 of 25 (100%) patients who had placement of a tracheoesophageal prosthesis achieved fluent speech. Median time to fluency was 63 days in the primary TEP group and 125 days in the secondary TEP group. CONCLUSION: There is an increased risk of PCF in patients undergoing primary TEP compared with secondary TEP following chemoradiation. No difference in acquisition of speech fluency was identified between the two groups. Patients undergoing primary TEP achieved fluent speech 62 days sooner than their secondary TEP counterparts. Sponsorships or competing interests that may be relevant to content are disclosed at the end of this article. © 2009 American Academy of Otolaryngology–Head and Neck Surgery Foundation. All rights reserved.
T
otal laryngectomy (TL) can be a life-saving procedure for patients with laryngeal cancer. Postoperative function, in particular swallowing and communication, has a significant impact on a patient’s quality of life following this ablative procedure. With respect to tracheoescophaeal (TE) speech, the goal is to achieve fluent TE voice and speech, which restores a patient’s ability to verbally communicate. To this end, TEP is the current standard of care and preferred method to restore speech communication. In 1980, Singer and Blom1 described their method for
voice restoration. They described a procedure of performing endoscopic tracheoesophageal puncture ultimately followed by placement of a one-way valve indwelling prosthesis. This technique has become widely accepted and is the preferred method for voice restoration.1-5 The procedure was initially described as a secondary procedure; however, over time this approach has evolved and it is now most commonly utilized in primary fashion at the time of TL. There is still controversy regarding the appropriate timing of the procedure, and any differences in voice outcome and complications related to this timing. Several series have demonstrated improved speech outcomes with primary vs secondary TEP.2-5 However, concern about increased risk for postoperative complications related to primary TEP makes secondary TEP favorable for some surgeons. A few series have included data on the impact of radiation therapy on voice outcomes and postoperative complications related to TEP.2,5 These data have suggested no significant impact on speech and limited impact on complications with respect to timing of TEP. However, to date, there are no data specific to TEP following concurrent chemoradiation. Because concurrent chemoradiation has become an increasingly utilized therapeutic regimen for advanced laryngeal cancer, we investigated complications and voice outcome in this population.
METHODS AND MATERIALS We conducted an institutional review board–approved (HUM00015899) retrospective review of a prospective cohort. Patients were identified from University of Michigan organ preservation protocols conducted between 1998 and 2005. All patients received induction chemotherapy followed by concurrent chemoradiation and salvage TL for persistent or recurrent disease. Patients who underwent an extended oncological resection beyond TL, including total
Received July 11, 2007; revised October 14, 2008; accepted October 14, 2008.
0194-5998/$36.00 © 2009 American Academy of Otolaryngology–Head and Neck Surgery Foundation. All rights reserved. doi:10.1016/j.otohns.2008.10.018
Emerick et al
Primary vs secondary tracheoesophageal . . .
pharyngolaryngectomy, TL with partial pharyngectomy, or TL with partial glossectomy, were included in this study. Those who received a total glossectomy, however, were excluded. Thirty patients were identified. All primary closures were performed in two layers in an interrupted fashion. Four patients required a total pharyngectomy. One was reconstructed with a free jejunal transfer, and the remaining procedures were reconstructed with a tubed fasciocutaneous free flap. In addition, all patients who could be closed primarily had a vascularized fascial flap placed over the pharyngeal closure. A cricopharyngeal myotomy was performed in all patients in which the cricopharyngeus was not resected. The decision to perform primary vs secondary TL was not influenced by patient factors such as intraoperative findings or complications, but rather exclusively by the intraoperative choice of the resecting surgeon. All four senior surgeons in this study have a comparable level of experience and routinely perform TL with primary and/or secondary TEP. Data were collected on each patient for underlying factors that could affect wound healing. Specifically, diabetes, hypoalbuminemia (preoperative albumin ⬍ 3.6 g/dL), hypothyroidism, systemic chronic disease, immunosuppression, coronary artery disease or peripheral vascular occlusive disease, anemia (hematocrit ⬍ 30% or by requirement for transfusions), thrombocytosis (platelets ⬎ 300 K/mm3), and extended oncological resection (total pharyngolaryngectomy, TL with partial pharyngectomy, or TL with partial glossectomy) were recorded. The record was reviewed for the following postoperative wound-healing complications: PCF, pharyngoesophageal stenosis, stomal stenosis, wound/ stomal breakdown (defined as localized wound dehiscence or tissue loss at the stoma without evidence of fistula involving the stoma), wound infection (defined as a localized cellulitis that resolved with a course of oral antibiotics), and leakage around the prosthesis. In the primary TEP group, the TEP prosthesis was placed at 3 to 4 weeks postoperatively, corresponding with the time the patient began oral intake. In patients with a fistula or
387
other complication, this procedure was delayed until the wound had sufficiently healed and there was no concern that digital manipulation of the prosthesis would impede continued healing. For the secondary TEP group, the puncture was attempted at 2 to 3 months postoperatively and the prosthesis placed 1 to 2 weeks following the puncture procedure. This timing allowed adequate healing of the pharyngeal closure to decrease the risk of injury with rigid esophagoscopy during prosthesis placement. Prosthesis placement and voice assessment were performed by a speech pathologist in all patients. The time to fluent speech acquisition was recorded for each patient. This value was calculated from the time of TL to the time of documented fluency by the speech pathologist. Wilcoxon rank-sum test was used to perform multivariate analysis of the comorbidities. In addition, univariate analysis using the Fisher exact test was also used to assess the individual comorbidities and differences in complications and speech fluency.
RESULTS Of the 30 patients who met the criteria for this study, 20 received primary TEP (17 men and three women) and 10 patients received secondary TEP (all men). Table 1 summarizes data and P values for each group regarding age, comorbidities, and extent of surgery. Univariate and multivariate analyses of these variables demonstrated no statistically significant difference between the groups and no correlation between the patient variables and postoperative complications. Postlaryngectomy-related complications for each TEP group are summarized in Table 2. The only statistically significant complication related to primary vs secondary TEP was PCF, with 50 percent (10/20) of patients who underwent primary TEP developing a PCF compared with 0 in the secondary TEP group (P ⫽ 0.006). All patients who developed fistulas in the primary TEP group did so in the first 2 months after
Table 1 Patient comorbidities and variables Patient variable
Primary TEP (N ⫽ 20)
Secondary TEP (N ⫽ 10)
P value
Mean age (y) Diabetes CAD or PVOD Hypothyroidism Hypoalbuminemia Systemic disease/infection Immunosuppression Anemia Thrombocytosis Extended oncological resection
62.2 15% 25% 25% 30% 15% 10% 65% 10% 40%
64.1 10% 10% 50% 10% 20% 0% 60% 20% 20%
⬎0.99 0.63 0.23 0.22 ⬎0.99 0.54 0.96 0.51 0.25
CAD, coronary artery disease; PVOD, peripheral vascular occlusive disease.
388
Otolaryngology–Head and Neck Surgery, Vol 140, No 3, March 2009
Table 2 Summary of incidence of complications
Complication PCF PE stenosis Wound/stomal Breakdown Wound infection TEP site leakage CP spasm Stomal stenosis
Table 4 Time to acquisition of fluent speech Median (d) Minimum Maximum
Primary TEP (N ⫽ 20)
Secondary TEP (N ⫽ 10)
P value
10 (50) 2 (10) 7 (35)
0 (0) 2 (2) 3 (40)
0.006* 0.31 0.49
1 1 1 0
0.59 0.46 0.46 1.0
3 1 1 0
(15) (5) (5) (0)
(10) (10) (10) (0)
PE, pharyngoesophageal; CP, cricopharyngeal. Values in parentheses are percentages of incidence for each group. *Statistically significant.
surgery, and more than half of these occurred in the first 2 weeks. Of the 10 fistulas that were observed, five healed in less than a month with conservative outpatient management, and another three healed within 3 months. Two patients had small, recurring fistulas that required surgical exploration and the creation of a controlled fistula. Each was then closed by secondary intention within 30 days. Additional analysis revealed that extended oncological resection was not a significant factor in fistula formation (P ⫽ 0.17). As far as speech fluency was concerned, all patients who had placement of a voice prosthesis achieved fluent speech (Table 3). Three patients in the primary TEP group and one in the secondary group who did not have a prosthesis placed were followed at an outside institution; therefore, voice data were not available. The fourth patient in the primary TEP group developed a rapid recurrence and was never fitted with a prosthesis. Median time to acquisition of fluency was much greater for the secondary TEP group (125 days) compared with the primary TEP group (63 days). As illustrated in Table 4, an even larger time difference was noted between primary TEP patients who did not develop fistulas and secondary TEP patients (48 vs 125 days, respectively).
Table 3 Acquisition of fluent speech
Primary TEP Secondary TEP All patients
Fluent speech
Prosthesis not placed
Fluency in those with prosthesis
16 (80) 9 (90) 25 (83)
4 (20) 1 (10) 5 (17)
16 (100) 9 (100) 25 (100)
Values in parentheses indicates percentage of entire group.
Primary TEP Primary TEP with fistula Primary TEP without fistula Secondary TEP
63 75
22 55
298 298
48
22
73
125
72
224
Most strikingly, however, even primary TEP patients who developed PCF still acquired fluency more quickly than secondary TEP patients as a group (75 vs 125 days, respectively).
DISCUSSION Voice restoration following TL is critical to the rehabilitation of these patients. Use on an electrolarynx, esophageal, and tracheoesophageal speech are all viable methods of voice restoration. However, since the introduction of alaryngea the Blom Singer prosthesis in 1980, tracheoesophageal speech is the most common modality used currently.1 This method was originally designed as a salvage procedure for those unable to utilize electrolarynx or esophageal speech. Over time application of tracheoesophageal speech has evolved, and it now is utilized as the first choice for voice rehabilitation in both the primary and secondary setting. In the past three decades there has been debate regarding the best timing for the procedure and whether the timing has an impact on postoperative complications and voice outcome. Because organ preservation protocols using neoadjuvant or concurrent chemoradiation therapy have become routine for advanced laryngeal cancer, we must consider the potential increased risk for TEP in the setting of salvage TL following concurrent chemoradiation. Previous studies on this subject have not been consistent in evaluating postoperative surgical complications. The Memorial Sloan Kettering group reported an overall surgical complication rate of 43 percent for primary TEP and 29 percent for secondary TEP; however, this finding was not statistically significant.2 Furthermore, a post hoc analysis of the impact of radiation did not demonstrate any difference in complications between the primary and secondary TEP groups. Similarly, Trudeau et al3 found timing of TEP to be independent of postoperative complications. A series of 38 patients reported by Wenig et al4 found a 10 percent PCF rate for primary TEP and none in the secondary group. A large series from Kao et al5 reported a similar 10 percent postoperative fistula rate in the primary TEP group, but the authors reported no data for the secondary TEP group. In our study, the two patient groups had similar comorbidities, and a multivariate analysis of patient’s comorbidi-
Emerick et al
Primary vs secondary tracheoesophageal . . .
ties specific to wound healing did not reveal any statistical differences. Therefore, we feel we can accurately assess the impact of timing of TEP on postsurgical complications. The only significant difference in complications noted was the incidence of PCF, 50 percent for the primary TEP group compared with zero percent for the secondary group. This complication can have significant morbidity. Fistulas can lead to prolonged hospitalizations, delay in oral intake, prolonged wound care, risk of catastrophic complications such as carotid blowout, and potential need for additional surgeries. In our practice, many of these issues have been addressed by utilizing free tissue transfer to reconstruct the resected pharyngeal segment or a free fascial flap to provide vascularized coverage of the pharyngeal closure and the major vessels.6 Although this technique has not significantly decreased the incidence of fistulas, it has limited the morbidity associated with the complication. The overall PCF rate for this series was 33 percent. Other series have reported similarly high PCF rates for salvage laryngectomy following chemoradiation.7,8 In this series, primary TEP was identified as a risk factor for PCF, and it may do so by creating a segment of devascularized or traumatized tissue between the inferior pharyngeal closure and the TEP site. This segment of compromised tissue may lead to a breakdown of the inferior suture line or necrosis of the intervening pharyngeal segment and ultimately PCF. If this etiology is assumed, it is unknown what the critical distance is to prevent vascular compromise to the intervening segment of pharyngeal mucosa. This series did not consider this distance; however, determining the critical distance may be an area of future investigation in an effort to better understand the etiology of fistula formation and ultimately to decrease the incidence of PCF in the primary TEP setting. The presence of a catheter to maintain patency of the TEP in the primary setting and additional digital manipulation associated with the care of a primary TEP raise the potential for increased stomal complications, such as dehiscence, narrowing, infection, and leakage from the TEP, in a recently created stoma. However, these complications were not an issue in this series; they were uncommon in both groups and not statistically different. Furthermore, complications specific to the puncture site, in particular leakage around the prosthesis, can also arise. The Memorial Sloan Kettering series did not reveal a significant difference between primary and secondary groups in this regard.2 However both series by Wenig and Kao reported a difference. Wenig4 noted a 10 percent incidence in leakage for the primary group and zero percent for the secondary group. To the contrary, Kao5 reported a 30 percent leakage complication for their secondary group and only a 2.8 percent incidence for the primary group. We did not find a difference in our patients. Much of the debate about primary vs secondary TEP has focused on voice-related issues. The first issue of great concern is voice acquisition. Most series report that greater
389
than 80 percent of patients achieve voice acquisition with acquisition among secondary TEP patients being lower.5,7,9,10 The second issue surrounding voice is the acquisition of fluent speech and the quality of the speech. Cheng et al2 found that a statistically significantly higher percentage of primary TEP patients achieved excellent voice quality compared with secondary TEP, 78 percent and 50 percent, respectively. Schultz and Harrison11 also found primary TEP to be a positive factor for predicting successful voice rehabilitation. Kao et al5 did not find a difference in voice outcome between primary and secondary TEP groups. From historical consideration of these data, it is difficult to compare variables such as quality of voice because each series has used different scales and definitions to assess these outcomes. With regard to voice issues, one limitation of our study is the lack of voice quality data. However, speech fluency data were available for all of the patients who had a prosthesis placed. Both groups achieved fluency in 100 percent of patients who had a prosthesis placed. Most series report a lower percentage of fluent patients, between 50 to 90 percent.2,4,5 A more formal voice evaluation would be necessary to identify small differences in the quality of speech production and its impact on quality of life between primary and secondary TEP. Early voice rehabilitation may lead to improved quality as suggested by some series. This improvement may be due to early use of the pharyngeal constrictors and mucosa for voice production prior to scarring and decreased tissue compliance that occurs over time in a surgical field compromised by chemoradiation. A third and final issue to consider regarding voice rehabilitation is time to voice acquisition. Inherently secondary TEP will more likely have a longer time to voice acquisition. In this study we found a long delay for the secondary group compared with the primary group (Table 4). The median time to voice acquisition for the primary TEP was 63 days, whereas the median for the secondary group was 125 days. This difference was even more pronounced when comparing uncomplicated primary TEP patients vs secondary TEP patients (48 vs 125 days, respectively). This prolonged delay in voice acquisition could have a significant impact on a patient’s quality of life. Other studies have reported similarly long times to voice acquisition for secondary TEP.4 Part of the delay is uncertainty regarding when it is safe to perform rigid esophagoscopy following TL. Most surgeons would favor longer rather than shorter time periods. Logistical factors such as surgical scheduling, adjuvant therapies, and patient distance from a medical center can delay the secondary procedure as well as the prosthesis placement. We did not record the time between prosthesis placement and voice acquisition; however, this parameter could be an important factor in explaining why voice acquisition was delayed longer in some patients than in others. One concern regarding primary TEP is the potential for postoperative PCF or other surgical complications leading to a delay in the acquisition of fluent speech. In this
390
Otolaryngology–Head and Neck Surgery, Vol 140, No 3, March 2009
population, we found that median time to voice acquisition for those patients who developed fistula was 75 days compared with 48 days for those without this complication. However, these values are still 7 weeks shorter than the median time for the secondary TEP group. Postoperative PCF did not prevent any patients from achieving fluent speech.
CONCLUSION This is the first study to compare the outcomes of primary vs secondary TEP in patients undergoing TL following concurrent chemoradiation therapy. Primary TEP is associated with an increased risk for developing PCF following TL in the post-chemoradiation patient. No additional complications were associated with primary TEP. There was no difference in speech fluency between the primary and secondary groups, and postoperative complications did not prevent anyone from achieving fluent speech. Finally, there is a large difference in the time to achieve speech fluency between these two groups. Even with consideration of postoperative PCF, primary TEP patients still acquire speech much sooner than their secondary counterparts. Both primary and secondary TEP should be considered safe options in the setting of salvage total laryngectomy following chemoradiation. Surgeons must weigh the potential increased risk for PCF in the primary setting against the significant delay in speech acquisition for secondary TEP.
AUTHOR INFORMATION From the Department of Otolaryngology–Head and Neck Surgery, University of Michigan (Drs Emerick, Bradford, Chepeha, and Wolf); the Department of Neurological Surgery, Vanderbilt University (Dr Tomycz); the Department of Speech and Language Pathology, University of Michigan (Dr Lyden); and the Department of Otolaryngology-Head and Neck Surgery, the Ohio State University Medical Center (Dr Teknos). Corresponding author: Theodoros N. Teknos, MD, Dept of Otolaryngology-Head and Neck Surgery, 456 W. 10th Ave, Suite 4A, Columbus, OH 43210. E-mail address: [email protected]. Paper presented as oral presentation at the Annual Meeting of the American Academy of Otolaryngology – Head and Neck Surgery Annual Meeting, Washington DC, September 16-19, 2007.
AUTHOR CONTRIBUTIONS Kevin Emerick, data management, authorship; Luke Tomycz, data management, authorship; Carol Bradford, editorial advice, data collection; Teresa Lyden, editorial advice, data collection; Douglas Chepeha, editorial advice, data collection; Gregory Wolf, editorial advice, data collection; Theodoros N. Teknos, data management, authorship, editorial advice.
FINANCIAL DISCLOSURES Kevin Emerick, none; Luke Tomycz, none; Carol Bradford, none; Teresa Lyden, none; Douglas Chepeha, principal investigator: KLS Martin grant (expired 1 year ago) for support of research coordinator; Gregory Wolf, consultant: IRX Therapeutics; Theodoros N. Teknos, none.
REFERENCES 1. Singer MI, Blom ED. An endoscopic technique for restoration of voice after laryngectomy. Ann Otol 1980;89:529 –33. 2. Cheng E, Ho M, Ganz C, et al. Outcomes of primary and secondary tracheoesophageal puncture: A 16-year retrospective analysis. ENT 2006;85:262–7. 3. Trudeau MD, Schuller DE, Hall DA. The timing of tracheoesophageal puncture for voice restoration: primary vs secondary. Head Neck Surg 1988;10:130 – 4. 4. Wenig BL, Levy J, Mullooly V, et al. Voice restoration following laryngectomy: the role of primary versus secondary tracheoesophageal puncture. Ann Otol Rhinol Laryngol 1989;98:70 –3. 5. Kao WW, Mohr RM, Kimmel CA, et al. The outcome and techniques of primary and secondary tracheoesophageal puncture. Arch Otolaryngol Head Neck Surg 1994;120:301–7. 6. Fung K, Teknos TN, Vandenberg CD, et al. Prevention of wound complications following salvage laryngectomy using free vascularized tissue. Head Neck 2007;29:425–30. 7. Weber RS, Burkey BA, Forastiere A, et al. Outcome of salvage total laryngectomy following organ preservation therapy—the Radiation Therapy Oncology Group trial 91-11. Arch Otolaryngol Head Neck Surg 2003;129:44 –9. 8. Sassler AM, Esclamado RM, Wolf GT. Surgery after organ preservation therapy—analysis of wound complications. Arch Otolaryngol Head Neck Surg 1995;121:162–5. 9. Izdebdski K, Reed CG, Ross JC, et al. Problems with tracheoesophageal fistula voice restoration in totally laryngectomized patients. Arch Otolaryngol Head Neck Surg 1994;120:840 –5. 10. Chone CT, Gripp FM, Spina AL, et al. Primary versus secondary tracheoesophageal puncture for speech rehabilitation in total laryngectomy: long-term results with indwelling voice prosthesis. Otolaryngol Head Neck Surg 2005;133:89 –93. 11. Shultz JR, Harrison J. Defining and predicting tracheoesophageal puncture success. Arch Otolaryngol Head Neck Surg 1992;118: 811– 6.
ORIGINAL RESEARCH—HEAD AND NECK SURGERY
Primary versus secondary tracheoesophageal puncture in salvage total laryngectomy following chemoradiation Kevin S. Emerick, MD, Luke Tomycz, MD, Carol R. Bradford, MD, Teresa H. Lyden, MA, CCC-SLP, Douglas B. Chepeha, MD, Gregory T. Wolf, MD, and Theodoros N. Teknos, MD, Ann Arbor, MI; Nashville, TN; and Columbus, OH OBJECTIVE: To compare the rate of postoperative wound-healing complications and voice fluency in primary vs secondary tracheoesophageal puncture (TEP) following chemoradiation. METHODS: Between 1998 and 2005, 30 patients underwent laryngectomy after chemoradiation therapy. Twenty patients underwent primary TEP and 10 patients underwent secondary TEP. Comorbidities, postoperative complications, speech fluency, and time to speech fluency were evaluated in each patient. RESULTS: Pharyngocutaneous fistula (PCF) occurred in 10 of 20 (50%) patients who underwent primary TEP and in 0 of 10 (0%) patients in the secondary TEP group (P ⬍ 0.05). Overall, 25 of 25 (100%) patients who had placement of a tracheoesophageal prosthesis achieved fluent speech. Median time to fluency was 63 days in the primary TEP group and 125 days in the secondary TEP group. CONCLUSION: There is an increased risk of PCF in patients undergoing primary TEP compared with secondary TEP following chemoradiation. No difference in acquisition of speech fluency was identified between the two groups. Patients undergoing primary TEP achieved fluent speech 62 days sooner than their secondary TEP counterparts. Sponsorships or competing interests that may be relevant to content are disclosed at the end of this article. © 2009 American Academy of Otolaryngology–Head and Neck Surgery Foundation. All rights reserved.
T
otal laryngectomy (TL) can be a life-saving procedure for patients with laryngeal cancer. Postoperative function, in particular swallowing and communication, has a significant impact on a patient’s quality of life following this ablative procedure. With respect to tracheoescophaeal (TE) speech, the goal is to achieve fluent TE voice and speech, which restores a patient’s ability to verbally communicate. To this end, TEP is the current standard of care and preferred method to restore speech communication. In 1980, Singer and Blom1 described their method for
voice restoration. They described a procedure of performing endoscopic tracheoesophageal puncture ultimately followed by placement of a one-way valve indwelling prosthesis. This technique has become widely accepted and is the preferred method for voice restoration.1-5 The procedure was initially described as a secondary procedure; however, over time this approach has evolved and it is now most commonly utilized in primary fashion at the time of TL. There is still controversy regarding the appropriate timing of the procedure, and any differences in voice outcome and complications related to this timing. Several series have demonstrated improved speech outcomes with primary vs secondary TEP.2-5 However, concern about increased risk for postoperative complications related to primary TEP makes secondary TEP favorable for some surgeons. A few series have included data on the impact of radiation therapy on voice outcomes and postoperative complications related to TEP.2,5 These data have suggested no significant impact on speech and limited impact on complications with respect to timing of TEP. However, to date, there are no data specific to TEP following concurrent chemoradiation. Because concurrent chemoradiation has become an increasingly utilized therapeutic regimen for advanced laryngeal cancer, we investigated complications and voice outcome in this population.
METHODS AND MATERIALS We conducted an institutional review board–approved (HUM00015899) retrospective review of a prospective cohort. Patients were identified from University of Michigan organ preservation protocols conducted between 1998 and 2005. All patients received induction chemotherapy followed by concurrent chemoradiation and salvage TL for persistent or recurrent disease. Patients who underwent an extended oncological resection beyond TL, including total
Received July 11, 2007; revised October 14, 2008; accepted October 14, 2008.
0194-5998/$36.00 © 2009 American Academy of Otolaryngology–Head and Neck Surgery Foundation. All rights reserved. doi:10.1016/j.otohns.2008.10.018
Emerick et al
Primary vs secondary tracheoesophageal . . .
pharyngolaryngectomy, TL with partial pharyngectomy, or TL with partial glossectomy, were included in this study. Those who received a total glossectomy, however, were excluded. Thirty patients were identified. All primary closures were performed in two layers in an interrupted fashion. Four patients required a total pharyngectomy. One was reconstructed with a free jejunal transfer, and the remaining procedures were reconstructed with a tubed fasciocutaneous free flap. In addition, all patients who could be closed primarily had a vascularized fascial flap placed over the pharyngeal closure. A cricopharyngeal myotomy was performed in all patients in which the cricopharyngeus was not resected. The decision to perform primary vs secondary TL was not influenced by patient factors such as intraoperative findings or complications, but rather exclusively by the intraoperative choice of the resecting surgeon. All four senior surgeons in this study have a comparable level of experience and routinely perform TL with primary and/or secondary TEP. Data were collected on each patient for underlying factors that could affect wound healing. Specifically, diabetes, hypoalbuminemia (preoperative albumin ⬍ 3.6 g/dL), hypothyroidism, systemic chronic disease, immunosuppression, coronary artery disease or peripheral vascular occlusive disease, anemia (hematocrit ⬍ 30% or by requirement for transfusions), thrombocytosis (platelets ⬎ 300 K/mm3), and extended oncological resection (total pharyngolaryngectomy, TL with partial pharyngectomy, or TL with partial glossectomy) were recorded. The record was reviewed for the following postoperative wound-healing complications: PCF, pharyngoesophageal stenosis, stomal stenosis, wound/ stomal breakdown (defined as localized wound dehiscence or tissue loss at the stoma without evidence of fistula involving the stoma), wound infection (defined as a localized cellulitis that resolved with a course of oral antibiotics), and leakage around the prosthesis. In the primary TEP group, the TEP prosthesis was placed at 3 to 4 weeks postoperatively, corresponding with the time the patient began oral intake. In patients with a fistula or
387
other complication, this procedure was delayed until the wound had sufficiently healed and there was no concern that digital manipulation of the prosthesis would impede continued healing. For the secondary TEP group, the puncture was attempted at 2 to 3 months postoperatively and the prosthesis placed 1 to 2 weeks following the puncture procedure. This timing allowed adequate healing of the pharyngeal closure to decrease the risk of injury with rigid esophagoscopy during prosthesis placement. Prosthesis placement and voice assessment were performed by a speech pathologist in all patients. The time to fluent speech acquisition was recorded for each patient. This value was calculated from the time of TL to the time of documented fluency by the speech pathologist. Wilcoxon rank-sum test was used to perform multivariate analysis of the comorbidities. In addition, univariate analysis using the Fisher exact test was also used to assess the individual comorbidities and differences in complications and speech fluency.
RESULTS Of the 30 patients who met the criteria for this study, 20 received primary TEP (17 men and three women) and 10 patients received secondary TEP (all men). Table 1 summarizes data and P values for each group regarding age, comorbidities, and extent of surgery. Univariate and multivariate analyses of these variables demonstrated no statistically significant difference between the groups and no correlation between the patient variables and postoperative complications. Postlaryngectomy-related complications for each TEP group are summarized in Table 2. The only statistically significant complication related to primary vs secondary TEP was PCF, with 50 percent (10/20) of patients who underwent primary TEP developing a PCF compared with 0 in the secondary TEP group (P ⫽ 0.006). All patients who developed fistulas in the primary TEP group did so in the first 2 months after
Table 1 Patient comorbidities and variables Patient variable
Primary TEP (N ⫽ 20)
Secondary TEP (N ⫽ 10)
P value
Mean age (y) Diabetes CAD or PVOD Hypothyroidism Hypoalbuminemia Systemic disease/infection Immunosuppression Anemia Thrombocytosis Extended oncological resection
62.2 15% 25% 25% 30% 15% 10% 65% 10% 40%
64.1 10% 10% 50% 10% 20% 0% 60% 20% 20%
⬎0.99 0.63 0.23 0.22 ⬎0.99 0.54 0.96 0.51 0.25
CAD, coronary artery disease; PVOD, peripheral vascular occlusive disease.
388
Otolaryngology–Head and Neck Surgery, Vol 140, No 3, March 2009
Table 2 Summary of incidence of complications
Complication PCF PE stenosis Wound/stomal Breakdown Wound infection TEP site leakage CP spasm Stomal stenosis
Table 4 Time to acquisition of fluent speech Median (d) Minimum Maximum
Primary TEP (N ⫽ 20)
Secondary TEP (N ⫽ 10)
P value
10 (50) 2 (10) 7 (35)
0 (0) 2 (2) 3 (40)
0.006* 0.31 0.49
1 1 1 0
0.59 0.46 0.46 1.0
3 1 1 0
(15) (5) (5) (0)
(10) (10) (10) (0)
PE, pharyngoesophageal; CP, cricopharyngeal. Values in parentheses are percentages of incidence for each group. *Statistically significant.
surgery, and more than half of these occurred in the first 2 weeks. Of the 10 fistulas that were observed, five healed in less than a month with conservative outpatient management, and another three healed within 3 months. Two patients had small, recurring fistulas that required surgical exploration and the creation of a controlled fistula. Each was then closed by secondary intention within 30 days. Additional analysis revealed that extended oncological resection was not a significant factor in fistula formation (P ⫽ 0.17). As far as speech fluency was concerned, all patients who had placement of a voice prosthesis achieved fluent speech (Table 3). Three patients in the primary TEP group and one in the secondary group who did not have a prosthesis placed were followed at an outside institution; therefore, voice data were not available. The fourth patient in the primary TEP group developed a rapid recurrence and was never fitted with a prosthesis. Median time to acquisition of fluency was much greater for the secondary TEP group (125 days) compared with the primary TEP group (63 days). As illustrated in Table 4, an even larger time difference was noted between primary TEP patients who did not develop fistulas and secondary TEP patients (48 vs 125 days, respectively).
Table 3 Acquisition of fluent speech
Primary TEP Secondary TEP All patients
Fluent speech
Prosthesis not placed
Fluency in those with prosthesis
16 (80) 9 (90) 25 (83)
4 (20) 1 (10) 5 (17)
16 (100) 9 (100) 25 (100)
Values in parentheses indicates percentage of entire group.
Primary TEP Primary TEP with fistula Primary TEP without fistula Secondary TEP
63 75
22 55
298 298
48
22
73
125
72
224
Most strikingly, however, even primary TEP patients who developed PCF still acquired fluency more quickly than secondary TEP patients as a group (75 vs 125 days, respectively).
DISCUSSION Voice restoration following TL is critical to the rehabilitation of these patients. Use on an electrolarynx, esophageal, and tracheoesophageal speech are all viable methods of voice restoration. However, since the introduction of alaryngea the Blom Singer prosthesis in 1980, tracheoesophageal speech is the most common modality used currently.1 This method was originally designed as a salvage procedure for those unable to utilize electrolarynx or esophageal speech. Over time application of tracheoesophageal speech has evolved, and it now is utilized as the first choice for voice rehabilitation in both the primary and secondary setting. In the past three decades there has been debate regarding the best timing for the procedure and whether the timing has an impact on postoperative complications and voice outcome. Because organ preservation protocols using neoadjuvant or concurrent chemoradiation therapy have become routine for advanced laryngeal cancer, we must consider the potential increased risk for TEP in the setting of salvage TL following concurrent chemoradiation. Previous studies on this subject have not been consistent in evaluating postoperative surgical complications. The Memorial Sloan Kettering group reported an overall surgical complication rate of 43 percent for primary TEP and 29 percent for secondary TEP; however, this finding was not statistically significant.2 Furthermore, a post hoc analysis of the impact of radiation did not demonstrate any difference in complications between the primary and secondary TEP groups. Similarly, Trudeau et al3 found timing of TEP to be independent of postoperative complications. A series of 38 patients reported by Wenig et al4 found a 10 percent PCF rate for primary TEP and none in the secondary group. A large series from Kao et al5 reported a similar 10 percent postoperative fistula rate in the primary TEP group, but the authors reported no data for the secondary TEP group. In our study, the two patient groups had similar comorbidities, and a multivariate analysis of patient’s comorbidi-
Emerick et al
Primary vs secondary tracheoesophageal . . .
ties specific to wound healing did not reveal any statistical differences. Therefore, we feel we can accurately assess the impact of timing of TEP on postsurgical complications. The only significant difference in complications noted was the incidence of PCF, 50 percent for the primary TEP group compared with zero percent for the secondary group. This complication can have significant morbidity. Fistulas can lead to prolonged hospitalizations, delay in oral intake, prolonged wound care, risk of catastrophic complications such as carotid blowout, and potential need for additional surgeries. In our practice, many of these issues have been addressed by utilizing free tissue transfer to reconstruct the resected pharyngeal segment or a free fascial flap to provide vascularized coverage of the pharyngeal closure and the major vessels.6 Although this technique has not significantly decreased the incidence of fistulas, it has limited the morbidity associated with the complication. The overall PCF rate for this series was 33 percent. Other series have reported similarly high PCF rates for salvage laryngectomy following chemoradiation.7,8 In this series, primary TEP was identified as a risk factor for PCF, and it may do so by creating a segment of devascularized or traumatized tissue between the inferior pharyngeal closure and the TEP site. This segment of compromised tissue may lead to a breakdown of the inferior suture line or necrosis of the intervening pharyngeal segment and ultimately PCF. If this etiology is assumed, it is unknown what the critical distance is to prevent vascular compromise to the intervening segment of pharyngeal mucosa. This series did not consider this distance; however, determining the critical distance may be an area of future investigation in an effort to better understand the etiology of fistula formation and ultimately to decrease the incidence of PCF in the primary TEP setting. The presence of a catheter to maintain patency of the TEP in the primary setting and additional digital manipulation associated with the care of a primary TEP raise the potential for increased stomal complications, such as dehiscence, narrowing, infection, and leakage from the TEP, in a recently created stoma. However, these complications were not an issue in this series; they were uncommon in both groups and not statistically different. Furthermore, complications specific to the puncture site, in particular leakage around the prosthesis, can also arise. The Memorial Sloan Kettering series did not reveal a significant difference between primary and secondary groups in this regard.2 However both series by Wenig and Kao reported a difference. Wenig4 noted a 10 percent incidence in leakage for the primary group and zero percent for the secondary group. To the contrary, Kao5 reported a 30 percent leakage complication for their secondary group and only a 2.8 percent incidence for the primary group. We did not find a difference in our patients. Much of the debate about primary vs secondary TEP has focused on voice-related issues. The first issue of great concern is voice acquisition. Most series report that greater
389
than 80 percent of patients achieve voice acquisition with acquisition among secondary TEP patients being lower.5,7,9,10 The second issue surrounding voice is the acquisition of fluent speech and the quality of the speech. Cheng et al2 found that a statistically significantly higher percentage of primary TEP patients achieved excellent voice quality compared with secondary TEP, 78 percent and 50 percent, respectively. Schultz and Harrison11 also found primary TEP to be a positive factor for predicting successful voice rehabilitation. Kao et al5 did not find a difference in voice outcome between primary and secondary TEP groups. From historical consideration of these data, it is difficult to compare variables such as quality of voice because each series has used different scales and definitions to assess these outcomes. With regard to voice issues, one limitation of our study is the lack of voice quality data. However, speech fluency data were available for all of the patients who had a prosthesis placed. Both groups achieved fluency in 100 percent of patients who had a prosthesis placed. Most series report a lower percentage of fluent patients, between 50 to 90 percent.2,4,5 A more formal voice evaluation would be necessary to identify small differences in the quality of speech production and its impact on quality of life between primary and secondary TEP. Early voice rehabilitation may lead to improved quality as suggested by some series. This improvement may be due to early use of the pharyngeal constrictors and mucosa for voice production prior to scarring and decreased tissue compliance that occurs over time in a surgical field compromised by chemoradiation. A third and final issue to consider regarding voice rehabilitation is time to voice acquisition. Inherently secondary TEP will more likely have a longer time to voice acquisition. In this study we found a long delay for the secondary group compared with the primary group (Table 4). The median time to voice acquisition for the primary TEP was 63 days, whereas the median for the secondary group was 125 days. This difference was even more pronounced when comparing uncomplicated primary TEP patients vs secondary TEP patients (48 vs 125 days, respectively). This prolonged delay in voice acquisition could have a significant impact on a patient’s quality of life. Other studies have reported similarly long times to voice acquisition for secondary TEP.4 Part of the delay is uncertainty regarding when it is safe to perform rigid esophagoscopy following TL. Most surgeons would favor longer rather than shorter time periods. Logistical factors such as surgical scheduling, adjuvant therapies, and patient distance from a medical center can delay the secondary procedure as well as the prosthesis placement. We did not record the time between prosthesis placement and voice acquisition; however, this parameter could be an important factor in explaining why voice acquisition was delayed longer in some patients than in others. One concern regarding primary TEP is the potential for postoperative PCF or other surgical complications leading to a delay in the acquisition of fluent speech. In this
390
Otolaryngology–Head and Neck Surgery, Vol 140, No 3, March 2009
population, we found that median time to voice acquisition for those patients who developed fistula was 75 days compared with 48 days for those without this complication. However, these values are still 7 weeks shorter than the median time for the secondary TEP group. Postoperative PCF did not prevent any patients from achieving fluent speech.
CONCLUSION This is the first study to compare the outcomes of primary vs secondary TEP in patients undergoing TL following concurrent chemoradiation therapy. Primary TEP is associated with an increased risk for developing PCF following TL in the post-chemoradiation patient. No additional complications were associated with primary TEP. There was no difference in speech fluency between the primary and secondary groups, and postoperative complications did not prevent anyone from achieving fluent speech. Finally, there is a large difference in the time to achieve speech fluency between these two groups. Even with consideration of postoperative PCF, primary TEP patients still acquire speech much sooner than their secondary counterparts. Both primary and secondary TEP should be considered safe options in the setting of salvage total laryngectomy following chemoradiation. Surgeons must weigh the potential increased risk for PCF in the primary setting against the significant delay in speech acquisition for secondary TEP.
AUTHOR INFORMATION From the Department of Otolaryngology–Head and Neck Surgery, University of Michigan (Drs Emerick, Bradford, Chepeha, and Wolf); the Department of Neurological Surgery, Vanderbilt University (Dr Tomycz); the Department of Speech and Language Pathology, University of Michigan (Dr Lyden); and the Department of Otolaryngology-Head and Neck Surgery, the Ohio State University Medical Center (Dr Teknos). Corresponding author: Theodoros N. Teknos, MD, Dept of Otolaryngology-Head and Neck Surgery, 456 W. 10th Ave, Suite 4A, Columbus, OH 43210. E-mail address: [email protected]. Paper presented as oral presentation at the Annual Meeting of the American Academy of Otolaryngology – Head and Neck Surgery Annual Meeting, Washington DC, September 16-19, 2007.
AUTHOR CONTRIBUTIONS Kevin Emerick, data management, authorship; Luke Tomycz, data management, authorship; Carol Bradford, editorial advice, data collection; Teresa Lyden, editorial advice, data collection; Douglas Chepeha, editorial advice, data collection; Gregory Wolf, editorial advice, data collection; Theodoros N. Teknos, data management, authorship, editorial advice.
FINANCIAL DISCLOSURES Kevin Emerick, none; Luke Tomycz, none; Carol Bradford, none; Teresa Lyden, none; Douglas Chepeha, principal investigator: KLS Martin grant (expired 1 year ago) for support of research coordinator; Gregory Wolf, consultant: IRX Therapeutics; Theodoros N. Teknos, none.
REFERENCES 1. Singer MI, Blom ED. An endoscopic technique for restoration of voice after laryngectomy. Ann Otol 1980;89:529 –33. 2. Cheng E, Ho M, Ganz C, et al. Outcomes of primary and secondary tracheoesophageal puncture: A 16-year retrospective analysis. ENT 2006;85:262–7. 3. Trudeau MD, Schuller DE, Hall DA. The timing of tracheoesophageal puncture for voice restoration: primary vs secondary. Head Neck Surg 1988;10:130 – 4. 4. Wenig BL, Levy J, Mullooly V, et al. Voice restoration following laryngectomy: the role of primary versus secondary tracheoesophageal puncture. Ann Otol Rhinol Laryngol 1989;98:70 –3. 5. Kao WW, Mohr RM, Kimmel CA, et al. The outcome and techniques of primary and secondary tracheoesophageal puncture. Arch Otolaryngol Head Neck Surg 1994;120:301–7. 6. Fung K, Teknos TN, Vandenberg CD, et al. Prevention of wound complications following salvage laryngectomy using free vascularized tissue. Head Neck 2007;29:425–30. 7. Weber RS, Burkey BA, Forastiere A, et al. Outcome of salvage total laryngectomy following organ preservation therapy—the Radiation Therapy Oncology Group trial 91-11. Arch Otolaryngol Head Neck Surg 2003;129:44 –9. 8. Sassler AM, Esclamado RM, Wolf GT. Surgery after organ preservation therapy—analysis of wound complications. Arch Otolaryngol Head Neck Surg 1995;121:162–5. 9. Izdebdski K, Reed CG, Ross JC, et al. Problems with tracheoesophageal fistula voice restoration in totally laryngectomized patients. Arch Otolaryngol Head Neck Surg 1994;120:840 –5. 10. Chone CT, Gripp FM, Spina AL, et al. Primary versus secondary tracheoesophageal puncture for speech rehabilitation in total laryngectomy: long-term results with indwelling voice prosthesis. Otolaryngol Head Neck Surg 2005;133:89 –93. 11. Shultz JR, Harrison J. Defining and predicting tracheoesophageal puncture success. Arch Otolaryngol Head Neck Surg 1992;118: 811– 6.