T2 Glottic Cancer Recurrence, Salvage, and Survival After Definitive Radiotherapy Dina
Howell-Burke, MD; Lester J. Peters, MD; Helmuth Goepfert, MD; Mary J. Oswald
\s=b\ The records of all patients with T2, NO squamous cell carcinoma of the true vocal cords treated with definitive radiotherapy at The University of Texas M. D. Anderson Cancer Center between 1970 and 1985 were analyzed to investigate treatment outcome, prognostic factors associated with tumor recurrence, and the potential impact of improved initial treatment on patient survival. There were 114 patients (male to female ratio, 13:1) with a median age of 62 years at presentation. All were treated with external beam irradiation to a modal dose of 70 Gy in 35 fractions over 7 weeks. The median field size was 25 cm2 and no elective treatment to the neck was routinely given. The crude recurrence rate after definitive radiotherapy was 32%. Of the 37 local regional failures, 32 were in the larynx only, 3 in the neck, and 2 in both the larynx and the neck. All patients who had recurrence after radiotherapy underwent salvage procedures, which increased the ultimate control rate above the clavicles to 94%. Overall and disease\x=req-\ specific survival rates at 5 years were 69% and 92%, respectively. Fifty patients died: 7 of laryngeal cancer, 2 of complications of salvage surgery, 13 of unrelated second
Accepted for publication March 30, 1990. From the Departments of Clinical Radiother-
apy (Drs Howell-Burke and Peters and Ms Oswald) and Head and Neck Surgery (Dr Goepfert), The University of Texas M. D. Anderson Cancer Center, Houston. Reprint requests to the Division of Radiotherapy (Box 97), University of Texas M. D. Anderson Cancer Center, 1515 Holcombe Blvd, Houston, TX 77030 (Dr Peters).
cancers, and 28 of other intercurrent disease. The prospects for improved survival through more effective initial treatment of this stage of glottic cancer are therefore very limited. Significant complications of radiotherapy occurred in only 4 patients (3.5%), and overall, 74% of patients retained a functional larynx. Analysis of a wide variety of patient-, tumor-, and treatment-related variables failed to identify any statistically significant prognostic factors. In the absence of reliable predictors of treatment failure, we continue to recommend definitive radiotherapy for patients with T2, NO glottic cancer. Since isolated neck recurrences are rare (2.6%), elective neck irradiation is not recommended. However, in patients undergoing salvage surgery for recurrence, postoperative radiotherapy to the stoma, neck, or both is advocated when the surgical pathologic examination reveals recurrent disease in the subglottis or thyroid cartilage, or extranodal disease in the neck. The treatment plan used in these circumstances is described. (Arch Otolaryngol Head Neck Surg.
1990;116:830-835)
Optimal
treatment of
laryngeal
demands both tumor eradication and preservation of laryn¬ geal function. Radiotherapy achieves this result in over 90% of patients with Tl glottic cancers, but tumor control declines progressively with increasing stage. Although some radiation failcancer
Downloaded From: http://archotol.jamanetwork.com/ by a University of Manitoba User on 06/07/2015
be salvaged by conservation total laryngectomy is usu¬ ally required. Methods to improve tu¬ mor control without sacrifice of the larynx include conservative surgical procedures, use of modified radiation dose fractionation schedules, and use of induction chemotherapy prior to definitive irradiation. It has been a traditional policy of this institution to treat most T2 glottic cancers by defin¬ itive radiotherapy. Rational applica¬ tion of possible alternative therapies requires analysis of causes and pat¬ terns of failure with conventional ra¬ diation therapy alone. This study was undertaken to (1) review the results of treatment of T2, NO glottic cancers by radiotherapy alone at The University of Texas M. D. Anderson Cancer Cen¬ ter, Houston, with particular refer¬ ence to the patient-, tumor-, and treat¬ ment-related parameters associated with tumor recurrence, and (2) exam¬ ine the potential impact of improved local disease control on patient sur¬ vival. ures can
surgery,
a
MATERIALS AND METHODS Patient Population Between
January 1970 and December
1985,114 patients with previously untreat¬ ed, biopsy-proved T2, NO squamous cell carcinoma of the glottis were treated with definitive radiotherapy at M. D. Anderson Cancer Center. The patient population con-
sisted of 101 men and 13 women with a me¬ dian age of 62 years (range, 38 to 76 years). The majority (82%) were identified as cig¬ arette smokers, with a median exposure of 40 pack-years; 15% had an uncertain smok¬ ing history; and only 3% were identified as nonsmokers. Performance status was not formally recorded in the majority of cases. However, most patients were in good to fair general condition. All the patients were staged clinically and radiologically. Most patients were re¬ ferred after endoscopy and biopsy had been performed. Direct endoscopie evaluation was limited to those patients who presented without a histologie diagnosis or for whom direct laryngoscopy was considered essen¬ tial for accurate delineation of disease. Ra¬ diologie examinations included conven¬ tional tomography, laryngography, and computed tomography, the latter being in¬ troduced for routine staging in 1982. All patients were evaluated by both a surgeon and a radiotherapist. The American Joint Committee on Can¬ cer criteria1 for stage T2 are (1) extension of disease from the vocal cords to the supra-
glottis
or
subglottis, (2) impaired
move¬
fixation, or (3) both. The distribution of patients by criteria for T2 staging is shown in Fig 1. ment of the vocal cord without
Treatment
All patients received definitive external beam irradiation (111 received cobalt 60 7 rays, and 3 received 6 MV x-rays) to total doses of 65 to 78 Gy in 6 to 8 weeks. The modal dose was 70 Gy in 35 fractions, and 88 patients (77%) received 70 ± 2.5 Gy. Ten patients in this series were treated with hyperfractionated irradiation to doses ranging from 74.4 to 76.8 Gy at 1.2 Gy per fraction twice daily. Forty-three patients were treated with ipsilateral fields only; the remainder were treated with parallel op¬ posed bilateral fields. The median field size was 25 cm2 (range, 16 to 100 cm2). Dosage was specified at the mid-vocal cord. Wedges were used only when disease was present in the posterior half of the vocal cord. All 37 patients whose primary radiother¬ apy failed underwent surgical salvage con¬ sisting of total laryngectomy in 30 patients (2 with simultaneous neck dissection), 1 hemilaryngectomy only, 3 neck dissections only, and 3 endoscopie laser ablations. Nine of the 37 patients whose initial treatment failed received additional radiotherapy for further recurrence after attempted surgi¬ cal salvage.
Follow-Up and End Points of Study
Follow-up information was obtained from clinic visits to M. D. Anderson Cancer
Supraglotiic Extension
Impaired Mobility
Subglottic Extension
Fig 1.—Distribution of patients with
T2
glottic carcinoma according
or by letters from referring physi¬ cians. All but 3 living patients were fol¬ lowed up from 36 to 192 months. Fifteen patients died without evidence of recur¬ rence above the clavicles before 36 months of follow-up. The primary end points of the study were to determine the probability of control of T2 glottic cancers by radiotherapy alone and after surgical salvage, to examine patient-, tumor-, and treatment-related parameters associated with an unsuccessful outcome, and to analyze survival and causes of death in this patient population, with particular reference to subsequent primary cancers outside the larynx. Criteria for distinguish¬ ing between lung métastases from the index laryngeal cancer and second primary lung cancers were (1) time greater than 5 years to development of lung lesion, and/or (2) different histopathologic features. Life-table analyses were performed us¬ ing the method of Berkson and Gage.2
Center,
RESULTS Local and Regional Disease Control
Of the 114 patients treated with de¬ finitive radiotherapy, local and/or re¬ gional failure occurred in 37: 32 (86% ) in the larynx, 2 in both the larynx and neck, and 3 in the neck only. Of the re¬ currences, 25 (68%) occurred within the first 2 years, 7 (19%) between 25 and 60 months, and 5 (14%) from 61 months to 15 years after treatment. Although late recurrences may have been second laryngeal primary can¬ cers, they were scored for this analysis as treatment failures. Surgical salvage was attempted in all 37 patients whose primary therapy
to
staging criteria.
failed; 30 of these patients received to¬ tal laryngectomy. The outcome of treatment of first recurrences by site of recurrence and treatment technique is shown in Table 1. Local-regional
disease control was achieved in 28 pa¬ tients (76%). Of the 9 second recur¬ rences, further treatment secured lo¬ cal-regional control in 5 (Table 2). Ul¬ timate local control through the time of death or last follow-up was there¬ fore obtained in all but 4 patients, and 84 patients (74%) retained a func¬ tional larynx. Distant Métastases
No patient whose primary cancer was controlled had a relapse at a dis¬ tant metastatic site. However, of the 37 patients who underwent salvage for local recurrence, regional recurrence, or both, 7 later developed distant mé¬ tastases. Four of these (3 with lung métastases and 1 with skin métastases to the scalp) survived 30 to 94 months after salvage surgery for metastatic disease and died of other causes. Survival
Fifty of the 114 patients died. A breakdown of causes of death is shown in Table 3. Absolute and disease-spe¬ cific survival is shown in Fig 2. Most patients died of causes unrelated to their index laryngeal cancer. Second Cancers
The cumulative probability of devel¬ oping a second cancer (other than skin
Downloaded From: http://archotol.jamanetwork.com/ by a University of Manitoba User on 06/07/2015
Table 1.—Outcome of Treatment of First Recurrences (37
Patients)*
Table 2.—Outcome of Treatment of Second Recurrences (9 Patients)*
Local-Regional Control by Treatment Modality
Site of Recurrence
Laryngectomy
Neck
and Neck Dissection
Dissection
Laryngectomy Only
Hemilaryn-
Radiotherapy
gectomy
Laser
Site of
Radiation
1/1
1/3t
Recurrence
Therapy
None
Stoma
2/3
0/1
Neck
1/3
23/28
Larynx
Local-Regional Control by Treatment Modality
Neck Dissection and
Surgery and
Neck
Larynx and neck
0/2
Larynx
*
Entries indicate number of patients whose first recurrence was controlled over the number of patients treated. fAlthough disease recurred in two patients treated by laser therapy, both recurrences were controlled by subsequent laser ablations.
cancer) after diagnosis of the glottic cancer is depicted in Fig 3. At 10 years,
the risk was 27%. The distribution of second cancers is shown in Table 4. Second cancers most frequently oc¬ curred in the lung and esophagus. The actuarial probability of death due to lung or esophageal cancer was 18.7% at 10 years, compared with 11.7% for the index glottic cancer. Two patients had second primary cancers, synchro¬ nous with their laryngeal cancer: one adenocarcinoma of esophagus and one small-cell lung cancer. Two additional patients (not included in the actuarial probability estimate of second prima¬ ries) had glottic cancer as their second and third primary cancers. Factors Influencing Outcome of
Primary Radiotherapy Patient-Related Variables.—No corre¬ lation was observed between the out¬ come of radiotherapy and age, sex, or amount of tobacco use. Treatment failed in 19 (35% ) of 55 patients below the median age of 62 years, compared with 18 (31 % ) of 59 patients 62 years or older. Treatment failed in 5 (39% ) of 13 women and in 32 (32%) of 101 men.
Cigarette consumption was quantitated in 84 patients. Of 31 patients with a median history of 40 pack-years or less, treatment failed in 11 (35% ). Of the 53 patients with a smoking history of greater than 40-pack years, treat¬ ment failed in 15 (28% ).
Histolo¬ Tumor Related Variables. gy.—All patients had squamous cell -
—
carcinomas. Histologie grade was specified in only 67 patients, of whom 12 had well-differentiated, 47 moder¬ ately differentiated, and 8 poorly dif¬ ferentiated tumors. Because of the in¬ complete data, no definite conclusions can be drawn regarding the probabil-
...
ity of treatment failure by tumor grade. However, the observed recur¬ rence rates for well-, moderately, and poorly differentiated tumors (3 of 12, 10 of 47, and 3 of 8, respectively) do not suggest any systematic relationship. Sites of Extension.—Pairwise com¬ parisons for the probability of recur¬ rence vs sites of tumor extension were tested separately for the anterior com¬ missure, contralateral true cord, ven¬ tricles, false cords, arytenoids, aryep¬
iglottic folds, infrahyoid epiglottis, and subglottis. Of these, only subglot¬ tic extension, as a single variable, was
associated with treatment failure. Subglottic extension was recorded in 87 patients, of whom 33 (38% ) suffered local recurrence compared with 4 (15%) of 27 without initial subglottic extension. The significance of subglot¬ tic extension as a prognostic factor
disappeared, however, in muitivariate analysis. Furthermore, review of the
failures among the 33 cases with initial subglottic extension who underwent salvage laryngectomy showed that only 7 patients had pathologic evi¬ dence of recurrence in the subglottis. The extent of initial subglottic exten¬ sion measured radiographically in these patients varied from 0.4 to 2.6 cm.
Cord
Mobility.—Impaired cord
mo¬
not associated with treat¬ ment failure: only 20 (27%) of 73 with
bility
was
impaired mobility recurred, compared
with 17 (41%) of 41 without demon¬ strated cord mobility. Criteria for T2 Staging.—The num¬ ber and combinations of criteria for T2 staging were also tested for prognostic significance (Fig 4). No combination of factors was significant by muitivariate
analysis.
Treatment-Related Variables.— Total
Downloaded From: http://archotol.jamanetwork.com/ by a University of Manitoba User on 06/07/2015
2/2
.
Entries indicate number of patients whose second recurrence was controlled over the number ot patients treated. ...
Table 3.—Causes of Death No. of Patients Index cancer Recurrence above clavicles Distant métastases
Surgical complications Second cancers Other intercurrent disease Total
4 3 2
13 28
50
Dose. —Total doses varied from 65 to 78 Gy. However, 88 of the 114 patients re¬ ceived tumor doses of 70 ± 2.5 Gy. This uniformity of dose prescription pre¬ cluded any meaningful dose-response
analysis.
Treatment Technique.—There was correlation between treatment out¬ come and size of radiation therapy fields. There were 19 (36% ) of 53 fail¬ ures with fields
Howell-Burke, MD; Lester J. Peters, MD; Helmuth Goepfert, MD; Mary J. Oswald
\s=b\ The records of all patients with T2, NO squamous cell carcinoma of the true vocal cords treated with definitive radiotherapy at The University of Texas M. D. Anderson Cancer Center between 1970 and 1985 were analyzed to investigate treatment outcome, prognostic factors associated with tumor recurrence, and the potential impact of improved initial treatment on patient survival. There were 114 patients (male to female ratio, 13:1) with a median age of 62 years at presentation. All were treated with external beam irradiation to a modal dose of 70 Gy in 35 fractions over 7 weeks. The median field size was 25 cm2 and no elective treatment to the neck was routinely given. The crude recurrence rate after definitive radiotherapy was 32%. Of the 37 local regional failures, 32 were in the larynx only, 3 in the neck, and 2 in both the larynx and the neck. All patients who had recurrence after radiotherapy underwent salvage procedures, which increased the ultimate control rate above the clavicles to 94%. Overall and disease\x=req-\ specific survival rates at 5 years were 69% and 92%, respectively. Fifty patients died: 7 of laryngeal cancer, 2 of complications of salvage surgery, 13 of unrelated second
Accepted for publication March 30, 1990. From the Departments of Clinical Radiother-
apy (Drs Howell-Burke and Peters and Ms Oswald) and Head and Neck Surgery (Dr Goepfert), The University of Texas M. D. Anderson Cancer Center, Houston. Reprint requests to the Division of Radiotherapy (Box 97), University of Texas M. D. Anderson Cancer Center, 1515 Holcombe Blvd, Houston, TX 77030 (Dr Peters).
cancers, and 28 of other intercurrent disease. The prospects for improved survival through more effective initial treatment of this stage of glottic cancer are therefore very limited. Significant complications of radiotherapy occurred in only 4 patients (3.5%), and overall, 74% of patients retained a functional larynx. Analysis of a wide variety of patient-, tumor-, and treatment-related variables failed to identify any statistically significant prognostic factors. In the absence of reliable predictors of treatment failure, we continue to recommend definitive radiotherapy for patients with T2, NO glottic cancer. Since isolated neck recurrences are rare (2.6%), elective neck irradiation is not recommended. However, in patients undergoing salvage surgery for recurrence, postoperative radiotherapy to the stoma, neck, or both is advocated when the surgical pathologic examination reveals recurrent disease in the subglottis or thyroid cartilage, or extranodal disease in the neck. The treatment plan used in these circumstances is described. (Arch Otolaryngol Head Neck Surg.
1990;116:830-835)
Optimal
treatment of
laryngeal
demands both tumor eradication and preservation of laryn¬ geal function. Radiotherapy achieves this result in over 90% of patients with Tl glottic cancers, but tumor control declines progressively with increasing stage. Although some radiation failcancer
Downloaded From: http://archotol.jamanetwork.com/ by a University of Manitoba User on 06/07/2015
be salvaged by conservation total laryngectomy is usu¬ ally required. Methods to improve tu¬ mor control without sacrifice of the larynx include conservative surgical procedures, use of modified radiation dose fractionation schedules, and use of induction chemotherapy prior to definitive irradiation. It has been a traditional policy of this institution to treat most T2 glottic cancers by defin¬ itive radiotherapy. Rational applica¬ tion of possible alternative therapies requires analysis of causes and pat¬ terns of failure with conventional ra¬ diation therapy alone. This study was undertaken to (1) review the results of treatment of T2, NO glottic cancers by radiotherapy alone at The University of Texas M. D. Anderson Cancer Cen¬ ter, Houston, with particular refer¬ ence to the patient-, tumor-, and treat¬ ment-related parameters associated with tumor recurrence, and (2) exam¬ ine the potential impact of improved local disease control on patient sur¬ vival. ures can
surgery,
a
MATERIALS AND METHODS Patient Population Between
January 1970 and December
1985,114 patients with previously untreat¬ ed, biopsy-proved T2, NO squamous cell carcinoma of the glottis were treated with definitive radiotherapy at M. D. Anderson Cancer Center. The patient population con-
sisted of 101 men and 13 women with a me¬ dian age of 62 years (range, 38 to 76 years). The majority (82%) were identified as cig¬ arette smokers, with a median exposure of 40 pack-years; 15% had an uncertain smok¬ ing history; and only 3% were identified as nonsmokers. Performance status was not formally recorded in the majority of cases. However, most patients were in good to fair general condition. All the patients were staged clinically and radiologically. Most patients were re¬ ferred after endoscopy and biopsy had been performed. Direct endoscopie evaluation was limited to those patients who presented without a histologie diagnosis or for whom direct laryngoscopy was considered essen¬ tial for accurate delineation of disease. Ra¬ diologie examinations included conven¬ tional tomography, laryngography, and computed tomography, the latter being in¬ troduced for routine staging in 1982. All patients were evaluated by both a surgeon and a radiotherapist. The American Joint Committee on Can¬ cer criteria1 for stage T2 are (1) extension of disease from the vocal cords to the supra-
glottis
or
subglottis, (2) impaired
move¬
fixation, or (3) both. The distribution of patients by criteria for T2 staging is shown in Fig 1. ment of the vocal cord without
Treatment
All patients received definitive external beam irradiation (111 received cobalt 60 7 rays, and 3 received 6 MV x-rays) to total doses of 65 to 78 Gy in 6 to 8 weeks. The modal dose was 70 Gy in 35 fractions, and 88 patients (77%) received 70 ± 2.5 Gy. Ten patients in this series were treated with hyperfractionated irradiation to doses ranging from 74.4 to 76.8 Gy at 1.2 Gy per fraction twice daily. Forty-three patients were treated with ipsilateral fields only; the remainder were treated with parallel op¬ posed bilateral fields. The median field size was 25 cm2 (range, 16 to 100 cm2). Dosage was specified at the mid-vocal cord. Wedges were used only when disease was present in the posterior half of the vocal cord. All 37 patients whose primary radiother¬ apy failed underwent surgical salvage con¬ sisting of total laryngectomy in 30 patients (2 with simultaneous neck dissection), 1 hemilaryngectomy only, 3 neck dissections only, and 3 endoscopie laser ablations. Nine of the 37 patients whose initial treatment failed received additional radiotherapy for further recurrence after attempted surgi¬ cal salvage.
Follow-Up and End Points of Study
Follow-up information was obtained from clinic visits to M. D. Anderson Cancer
Supraglotiic Extension
Impaired Mobility
Subglottic Extension
Fig 1.—Distribution of patients with
T2
glottic carcinoma according
or by letters from referring physi¬ cians. All but 3 living patients were fol¬ lowed up from 36 to 192 months. Fifteen patients died without evidence of recur¬ rence above the clavicles before 36 months of follow-up. The primary end points of the study were to determine the probability of control of T2 glottic cancers by radiotherapy alone and after surgical salvage, to examine patient-, tumor-, and treatment-related parameters associated with an unsuccessful outcome, and to analyze survival and causes of death in this patient population, with particular reference to subsequent primary cancers outside the larynx. Criteria for distinguish¬ ing between lung métastases from the index laryngeal cancer and second primary lung cancers were (1) time greater than 5 years to development of lung lesion, and/or (2) different histopathologic features. Life-table analyses were performed us¬ ing the method of Berkson and Gage.2
Center,
RESULTS Local and Regional Disease Control
Of the 114 patients treated with de¬ finitive radiotherapy, local and/or re¬ gional failure occurred in 37: 32 (86% ) in the larynx, 2 in both the larynx and neck, and 3 in the neck only. Of the re¬ currences, 25 (68%) occurred within the first 2 years, 7 (19%) between 25 and 60 months, and 5 (14%) from 61 months to 15 years after treatment. Although late recurrences may have been second laryngeal primary can¬ cers, they were scored for this analysis as treatment failures. Surgical salvage was attempted in all 37 patients whose primary therapy
to
staging criteria.
failed; 30 of these patients received to¬ tal laryngectomy. The outcome of treatment of first recurrences by site of recurrence and treatment technique is shown in Table 1. Local-regional
disease control was achieved in 28 pa¬ tients (76%). Of the 9 second recur¬ rences, further treatment secured lo¬ cal-regional control in 5 (Table 2). Ul¬ timate local control through the time of death or last follow-up was there¬ fore obtained in all but 4 patients, and 84 patients (74%) retained a func¬ tional larynx. Distant Métastases
No patient whose primary cancer was controlled had a relapse at a dis¬ tant metastatic site. However, of the 37 patients who underwent salvage for local recurrence, regional recurrence, or both, 7 later developed distant mé¬ tastases. Four of these (3 with lung métastases and 1 with skin métastases to the scalp) survived 30 to 94 months after salvage surgery for metastatic disease and died of other causes. Survival
Fifty of the 114 patients died. A breakdown of causes of death is shown in Table 3. Absolute and disease-spe¬ cific survival is shown in Fig 2. Most patients died of causes unrelated to their index laryngeal cancer. Second Cancers
The cumulative probability of devel¬ oping a second cancer (other than skin
Downloaded From: http://archotol.jamanetwork.com/ by a University of Manitoba User on 06/07/2015
Table 1.—Outcome of Treatment of First Recurrences (37
Patients)*
Table 2.—Outcome of Treatment of Second Recurrences (9 Patients)*
Local-Regional Control by Treatment Modality
Site of Recurrence
Laryngectomy
Neck
and Neck Dissection
Dissection
Laryngectomy Only
Hemilaryn-
Radiotherapy
gectomy
Laser
Site of
Radiation
1/1
1/3t
Recurrence
Therapy
None
Stoma
2/3
0/1
Neck
1/3
23/28
Larynx
Local-Regional Control by Treatment Modality
Neck Dissection and
Surgery and
Neck
Larynx and neck
0/2
Larynx
*
Entries indicate number of patients whose first recurrence was controlled over the number of patients treated. fAlthough disease recurred in two patients treated by laser therapy, both recurrences were controlled by subsequent laser ablations.
cancer) after diagnosis of the glottic cancer is depicted in Fig 3. At 10 years,
the risk was 27%. The distribution of second cancers is shown in Table 4. Second cancers most frequently oc¬ curred in the lung and esophagus. The actuarial probability of death due to lung or esophageal cancer was 18.7% at 10 years, compared with 11.7% for the index glottic cancer. Two patients had second primary cancers, synchro¬ nous with their laryngeal cancer: one adenocarcinoma of esophagus and one small-cell lung cancer. Two additional patients (not included in the actuarial probability estimate of second prima¬ ries) had glottic cancer as their second and third primary cancers. Factors Influencing Outcome of
Primary Radiotherapy Patient-Related Variables.—No corre¬ lation was observed between the out¬ come of radiotherapy and age, sex, or amount of tobacco use. Treatment failed in 19 (35% ) of 55 patients below the median age of 62 years, compared with 18 (31 % ) of 59 patients 62 years or older. Treatment failed in 5 (39% ) of 13 women and in 32 (32%) of 101 men.
Cigarette consumption was quantitated in 84 patients. Of 31 patients with a median history of 40 pack-years or less, treatment failed in 11 (35% ). Of the 53 patients with a smoking history of greater than 40-pack years, treat¬ ment failed in 15 (28% ).
Histolo¬ Tumor Related Variables. gy.—All patients had squamous cell -
—
carcinomas. Histologie grade was specified in only 67 patients, of whom 12 had well-differentiated, 47 moder¬ ately differentiated, and 8 poorly dif¬ ferentiated tumors. Because of the in¬ complete data, no definite conclusions can be drawn regarding the probabil-
...
ity of treatment failure by tumor grade. However, the observed recur¬ rence rates for well-, moderately, and poorly differentiated tumors (3 of 12, 10 of 47, and 3 of 8, respectively) do not suggest any systematic relationship. Sites of Extension.—Pairwise com¬ parisons for the probability of recur¬ rence vs sites of tumor extension were tested separately for the anterior com¬ missure, contralateral true cord, ven¬ tricles, false cords, arytenoids, aryep¬
iglottic folds, infrahyoid epiglottis, and subglottis. Of these, only subglot¬ tic extension, as a single variable, was
associated with treatment failure. Subglottic extension was recorded in 87 patients, of whom 33 (38% ) suffered local recurrence compared with 4 (15%) of 27 without initial subglottic extension. The significance of subglot¬ tic extension as a prognostic factor
disappeared, however, in muitivariate analysis. Furthermore, review of the
failures among the 33 cases with initial subglottic extension who underwent salvage laryngectomy showed that only 7 patients had pathologic evi¬ dence of recurrence in the subglottis. The extent of initial subglottic exten¬ sion measured radiographically in these patients varied from 0.4 to 2.6 cm.
Cord
Mobility.—Impaired cord
mo¬
not associated with treat¬ ment failure: only 20 (27%) of 73 with
bility
was
impaired mobility recurred, compared
with 17 (41%) of 41 without demon¬ strated cord mobility. Criteria for T2 Staging.—The num¬ ber and combinations of criteria for T2 staging were also tested for prognostic significance (Fig 4). No combination of factors was significant by muitivariate
analysis.
Treatment-Related Variables.— Total
Downloaded From: http://archotol.jamanetwork.com/ by a University of Manitoba User on 06/07/2015
2/2
.
Entries indicate number of patients whose second recurrence was controlled over the number ot patients treated. ...
Table 3.—Causes of Death No. of Patients Index cancer Recurrence above clavicles Distant métastases
Surgical complications Second cancers Other intercurrent disease Total
4 3 2
13 28
50
Dose. —Total doses varied from 65 to 78 Gy. However, 88 of the 114 patients re¬ ceived tumor doses of 70 ± 2.5 Gy. This uniformity of dose prescription pre¬ cluded any meaningful dose-response
analysis.
Treatment Technique.—There was correlation between treatment out¬ come and size of radiation therapy fields. There were 19 (36% ) of 53 fail¬ ures with fields
