Investigation of patients with apparently inoperable lung cancer

O pa tients pre sent ing with bron cho genic car ci noma, roughly onethird have stage I to II dis ease that ap pears pos si bly ame na ble to sur gi cal cure, onethird have dis tant me tas ta ses, and onethird have lo cally ad vanced dis ease that is po ten tially re secta ble (1). Of the lat ter, as few as 15% are can di dates for sur gi cal cure. Clini cal and ra dio logi cal stag ing (in clud ing his tory, physi cal ex ami na tion, chest ra di og ra phy and com puted to mo gra phy [CT] scan) are not spe cific enough to rule out the pos si bil ity of re sec tion for cure, es pe cially when the cri te ria for in op er abil ity in clude pre sumed posi tive me di as ti nal (N2 or N3) nodes (Ta bles 1,2) (2). We pres ent our ex pe ri ence with 30 pa tients with bron cho genic can cer whose ini tial workup sug gested ad vanced lo cal or re gional dis ease. Ten pa tients, who were ul ti mately found to be can di dates for sur gi cal re sec tion, formed the ba sis of this study.


In ves ti ga tion des pa tients at te ints d'un can cer du pou mon ap paremment inopé rable
RÉS UMÉ : Pen dant une pé ri ode de trois ans, 30 pa tients ont été ad res sés à un groupe mul ti dis ci pli naire spé ci al isé en can cer du pou mon afin de subir une ra dio thé ra pie pal lia tive pour un can cer bron chique prés umé non résé ca ble d'a près les résul tats ra dio graphiques.Une in ves ti ga tion plus pous sée, y com pris une clas si fi ca tion par sta des chi rur gi caux, a démon tré que 10 pa tients pou vaient subir une ré sec tion.Neuf pa tients sont vi vants et en rémis sion après un suivi médian de 25,8 mois.Ces résul tats démontrent que la radiogra phie et la to moden si tométrie tho ra cique pour raient man quer de spé ci fic ité pour re je ter la pos si bil ité d'un traitement chi rur gi cal.pa tients with pri mary nonsmall cell can cer (NSCC) of the lung should not be ruled out for cura tive re sec tion sim ply on the ba sis of CT scan.The role of sur gi cal stag ing in such cases is re viewed in light of these find ings.

POPULATION STUDIED
The pa tients who are de scribed here were seen over a three-year pe riod by three of the authors at the Lung Can cer Clinic at the WW Cross, a mul ti dis ci plin ary group as so ci ated with the Uni ver sity of Al berta.The group as sesses 500 to 600 cases each year.Ap proxi mately 20% of pa tients are re ferred fol low ing sur gery, 20% fol low ing posi tive me di asti noscopy and the ma jor ity of the re main der have docu mented me tas ta -ses, physio logi cal status pre clud ing sur gery or docu mented lo cal ex ten sion be yond the limit of sur gery.Very few patients (fewer than 5%) are re ferred for pal lia tive ra dia tion on the ba sis of pre sumed, but not proven, lo cally ad vanced disease.This se lect sub group is con sid ered for ag gres sive staging on an in di vid ual ba sis to de fine the pos si bil ity of cura tive re sec tion or en try into clini cal tri als.

INTERVENTIONS
Pa tients who ap pear to have the po ten tial for sur gi cal cure are first evalu ated to de ter mine their abil ity to un dergo complete re sec tion.Physio logi cal as sess ment in cludes stair climb ing, pul mo nary func tion tests and, where needed, quanti ta tive split lung func tions.CT of the chest and up per ab domen is ob tained if not al ready done.With ele va tion of se rum al ka line phos phatase con cen tra tion or clini cal evi dence sugges tive of bony me tas ta ses, bone scans are per formed.Head CTs are per formed based on symp to matol ogy.If no evi dence of meta static dis ease is found, and the pa tient is deemed able to un dergo re sec tion, me di asti noscopy is per formed.Four sta tions are rou tinely sam pled.If me di asti noscopy is positive, the pa tient is given the op tion of en roll ing in clini cal trials of in duc tion ther apy.Tho ra coscopy is per formed in the pres ence of cy to logi cally nega tive pleu ral ef fu sion, and is con sid ered if there is sug ges tion of me di as ti nal or ver te bral in va sion.If tho ra coscopy is nega tive, the pa tient un der goes tho ra cotomy with the goal of com plete re sec tion, as de fined by Moun tain (3).The fi nal stage is clas si fied ac cord ing to the TNM (tu mour, node, me tas ta sis) method de scribed by Moun tain (4) in 1986 (Ta bles 1 and 2) (5,6).A tumour that is 3.0 cm or less in greatest dimension, surrounded by lung or visceral pleura, and without evidence of invasion proximal to a lobar bronchus at bronchoscopy T2 A tumour more than 3.0 cm in greatest dimension, or a tumour of any size that either invades the visceral pleura or has associated atelectasis or obstructive pneumonitis extending to the hilar region.At bronchoscopy, the proximal extent of demonstrable tumour must be within a lobar bronchus or at least 2.0 cm distal to the carina.Any associated atelectasis or obstructive pneumonitis must involve less than an entire lung T3 A tumour of any size with direct extension into the chest wall (including superior sulcus tumours), diaphragm, or the mediastinal pleura or pericardium without involving the heart, great vessels, trachea, esophagus or vertebral body, or a tumour in the main bronchus within 2 cm of the carina without involving the carina  MAIN RESULTS Be tween July 1991 and June 1994 ap proxi mately 1300 pa tients were re ferred to the group with new lung can cers.Thirty pa tients, re ferred for pal lia tive ra dio ther apy based on ra dio graphic evi dence of lo cal or re gional ad vanced dis ease, were fur ther in ves ti gated and found to have nega tive metastatic work-up.These pa tients un der went sur gi cal stag ing.Twenty pa tients had posi tive me di asti noscopy and were random ized to in duc tion pro to cols.
Ten pa tients (six male, four fe male) (Ta ble 3) with biopsy-proven squa mous cell can cer were found to be po tentially re secta ble.CT in di ca tions of ex ten sive lo cal or re gional dis ease in cluded evi dence of en larged N2 nodes (7), me di asti nal in va sion (3), ef fu sion (1) and pos si ble ad re nal me tas tases (1) (Ta ble 4).In ad di tion four pa tients were thought to have car dio pul mon ary re serves in com pati ble with sur gery.Eight pa tients had right-sided and two had left-sided le sions.Cer vi cal me di asti noscopy was nega tive in all pa tients.The pa tients with left-sided le sions (#6, #8; Ta ble 4) un der went an te rior me di astinotomy that re vealed the pres ence of small mo bile nodes with only mi cro scopic in volve ment in one patient (#8).A fine nee dle as pi rate of the sus pected ad re nal gland le sion re vealed no evi dence of meta static dis ease.All pa tients were able to un dergo com plete re sec tion and all were alive at a mean follow-up of 25.8 months (Ta ble 4).One patient (#3) pre sented with an un tapped ef fu sion and mar ginal pul mo nary func tion; tho ra cen te sis and tho ra coscopy did not re veal evi dence of ma lig nancy and he un der went bi lobectomy.Pa thol ogy re vealed poorly dif fer en ti ated squa mous cell car ci noma and posi tive pos te rior sub cari nal ade no pa thy.He de clined ad ju vant ther apy, and un for tu nately, at 11 months' follow-up, he pre sented with liver me tas ta ses.The other nine pa tients, all of whom had well-differentiated squamous cell can cer, are disease-free at the time of writ ing.

DISCUSSION
We have pre sented a se ries of pa tients who were thought, on the ba sis of tho racic im ag ing, to be in op er able be cause of in tra tho racic spread and were re ferred for ra dio ther apy.How ever, sur gi cal stag ing re vealed that 10 were, in fact, resecta ble.This se ries dem on strates that ra dio logi cal evi dence of un re secta ble lo cal or re gional dis ease may not pre clude a chance of com plete re sec tion and pos si ble cure.A pro por tion of such pa tients, who are physio logi cally ca pa ble of withstand ing cura tive re sec tion and do not have ex tra tho racic metas ta ses, pres ent with chest x-ray or CT evi dence of one or more of the fol low ing: posi tive N2 nodal me tas ta ses; in vasion of the heart, great ves sels, esopha gus and/or ver te brae; and ma lig nant pleu ral ef fu sion or pleu ral seed ing.A va ri ety of ap proaches are avail able to con firm un re secta bil ity, and should be con sid ered to 'r atio nally man age' these pa tients (6).The fact that all the pa tients who were found to be op erative can di dates in this study had squa mous cell can cer probably re flects the greater ten dancy of squa mous cell can cer to spread lo cally rather than sys temi cally, com pared with other forms of NSCC (8).In fact, the prin ci ples de scribed ap ply to all forms of NSCC.
Ra dio logi cal fea tures that im ply lo cal or re gional dis ease pre clud ing re sec tion in clude evi dence of me di as ti nal ade nopa thy, me di as ti nal or ver te bral in va sion and pleu ral ef fu sion.The po ten tial role of sur gi cal stag ing in each of these set tings will be dis cussed in turn.
Ini tial as sess ment of me di as ti nal lymph nodes is by chest x-ray and chest CT.Pa tients with me di as ti nal ade no pa thy noted on plain chest x-ray have only a 9% three-year sur vival (8).The pres ence of me di as ti nal lym pha de no pa thy greater than 1 cm in di ame ter, while usu ally in dica tive of me tas ta ses, does not ap pear to be spe cific enough to rule out the po ten tial for com plete re secta bil ity, and cer tainly patho logi cal con firma tion is re quired (2,6,7).This can be ob tained pretho racotomy by me di asti noscopy, me di astinotomy or tho ra coscopy.Me di asti noscopy is safe (less than 2% mor bidity) and re duces nega tive ex plo ra tion rates to less than 5%    L Left; R Right (10).Docu men ta tion of posi tive N2 nodes does not uniformly pre clude re sec tion.Ap proxi mately 10% of these patients have low, ip si lat eral, single-station in volve ment with no fixa tion or ex tra cap su lar ex ten sion and may bene fit from tho ra cotomy (11).Pa tients with left-sided le sions who have nega tive cer vi cal me di asti noscopy but mi cro scopic, although re secta ble, me tas ta ses to N2 nodes in the aor to pulmon ary win dow alone may also be con sid ered for cura tive re sec tion be cause the five-year sur vival rate af ter com plete re sec tion in the pres ence of only mi cro scopic subaor tic nodal dis ease ap proaches that of sur vival af ter re sec tion of N1 disease (12).Thus, even with docu mented N2 dis ease, some patients with NSCC may be re secta ble, as long as the prin ci ples laid out by Put nam (2) are fol lowed (Ta ble 5).Me di asti noscopy may also be con sid ered if there is other evi dence of lo cal dis ease ex ten sion, such as chest wall in vasion, where docu mented me di as ti nal spread would usu ally pre clude tho ra cotomy.Ad di tion ally, while it may be more cost ef fec tive to avoid me di asti noscopy when CT does not re veal en larged (greater than 1 cm) me di as ti nal nodes, me diasti noscopy may still be ap pro pri ate in pa tients whose general medi cal con di tion puts them at sig nifi cantly in creased risk for sur gery (13), be cause up to 15% of lymph nodes that are 'no rmal' by CT con tain me tas ta ses, and me di asti noscopy may avoid a 'nonther ape utic' tho ra cotomy (6,7).A simi lar situa tion oc curs when the pri mary is sus pected to be small cell can cer, where me di asti noscopic con fir ma tion would also pre clude tho ra cotomy (6,13).
Apart from me di as ti nal ade no pa thy, other CT mani fes tations of lo cal or re gional spread that im ply 'u nr esect abi lity' in clude the sug ges tion of ver te bral or me di as ti nal in va sion.In the ab sence of ob vi ous in va sion or en cir clem ent there may be only in di rect evi dence of abut ment or loss of the fat plane be tween the ma lig nancy and ad join ing me di as ti nal structures.In the pres ence of these in de ter mi nant find ings, resecta bil ity is pos si ble in up to 82% of cases (14).More spe cifi cally, im por tant fea tures to evalu ate are less than 3 cm of con tact with me di as ti nal struc tures, less than 90° of contact with the aorta and the pres ence of a me di as ti nal fat plane.Re sec tion is pos si ble in up to 97% of cases where any one of these is pres ent (14).Other op tions that should be con sid ered in clude the use of eso pha go scopy, bone scans, mag netic reso nance im ag ing and an gi ogra phy.Not only does pretho racotomy tho ra coscopy of fer a means of con firm ing in va sion, but it should also proba bly be con sid ered man da tory in the pres ence of pleu ral ef fu sion.The re secta bil ity rate, even in the pres ence of cy to logi cally nega tive fluid, may be as low as 5.5% (6,15).How ever, ul ti mately fi nal stag ing may re quire tho ra cotomy (6).
The bulk of the above dis cus sion has con cen trated on the role and ac cu racy of stag ing in re gard to the pos si bil ity of sur gi cal ex tir pa tion.Even when pretho ra cotomy sur gi cal stag ing con fims ex ten sive dis ease pre clud ing cura tive re section, ac cu rate stag ing may al low en try into in duc tion pro tocols (16,17).Ma jor re sponse rates in stage III bron cho genic can cers of up to 77% have been de scribed, with five-year survival in com plete re spond ers who un dergo com plete re section equal ling that of mediastinoscopy-negative pa tients (17).

CONCLUSION
Pa tients with NSCC of the lung may pres ent with sug gestive ra dio graphic evi dence of in op er abil ity due to lo cally advanced dis ease.In deed, the ma jor ity have in tra-or ex tra tho racic spread that does pre clude cura tive re sec tion.How ever, be cause the speci fic ity of such evi dence is too low to rule any pa tient out on that ba sis alone, his to logi cal con firma tion is re quired.Moreo ver, ac cu rate stag ing al lows bet ter 'pr ot oc ol iz ation' be cause some of these pa tients may bene fit from in duc tion tri als.Us ing the ad mit tedly labour-intensive ap proach de scribed, 10 pa tients who were re ferred as 'i no perable' can di dates for ra dio ther apy were able to un dergo complete re sec tion, with nine be ing free of dis ease at a mean follow-up of 25.8 months.Can Respir J Vol 3 No 5 Sep tem ber/Oc to ber 1996 Karmy-Jones et al TABLE 5 Determinants of resectibility for patients with documented or suspected N2 disease 1.The patient must be physiologically capable of withstanding the planned pulmonary resection 2. The planned resection must encompass the primary tumour and all draining nodal beds 3.There are no distant metastases, eg, the disease is confined to the ipsilateral hemithorax 4. If N2 disease is suspected, an attempt should be made to confirm histologically the presence of N2 disease before thoracotomy and, if histology is positive, the patient should be entered into prospective randomized trials.If histology is negative, the patient should proceed with thoracotomy, pulmonary resection and mediastinal lymph node dissection 5.When there is doubt as to the presence of N2 disease, the patient should be given the benefit of staging and surgical resection.Patients with obstructive pneumonia may have evidence of enlarged hilar or mediastinal adenopathy.Mediastinoscopy should be performed in these patients.(See 3 above) 6.If N2 disease is documented after surgical resection, the patient should be entered into prospective randomized trials 7. Patients with T3N2 disease confirmed histologically (by mediastinoscopy, mediastinotomy or video-assisted thorascopic surgery) should not routinely have surgical resection because postresection survival is poor.Resection may be performed for palliation

310 Can Respir J Vol 3
No 5 Sep tem ber/Oc to ber 1996Karmy-Jones et al

: Nodal in volve ment N0
T4A tumour of any size with invasion of the mediastinum or involving heart, great vessels, trachea, esophagus or vertebral body or carina, or presence of malignant pleural effusion N No demonstrable metastasis to regional lymph nodes N1 Metastasis to lymph nodes in the peribronchial or the ipsilateral hilar region, or both, including direct extension N2 Metastasis to ipsilateral mediastinal lymph nodes and subcarinal lymph nodes N3 Metastasis to contralateral mediastinal lymph nodes, contralateral hilar lymph nodes, or ipsilateral or contralateral scalene or supraclavicular lymph nodes M:

par ently in op er able lung can cer
Can Respir J Vol 3 No 5 Sep tem ber/Oc to ber 1996 311 Ap