An experimental drug cocktail that killed a record 99 percent of TB bacteria has been unveiled by researchers.
It may also, crucially, be compatible with AIDS medicine.
While TB claims a life every 20 seconds, mainly in poor countries, more people are developing resistance to the existing arsenal of drugs that can take two years to work and often interacts badly with HIV antiretrovirals.
The sputum test is an early indicator of how quickly a drug works, but further clinical tests are required to confirm that it does in fact cure TB, a lung disease that spreads through coughing and sneezing.
Spigelman and the research team said they were confident that further trials would confirm early, encouraging results for the three-drug combination called PaMZ.
The first tests with 85 patients suggested the drug could potentially cure people with ordinary TB and some forms of multi-drug resistance (MDR) "in as little as four months", Spigelman told journalists in a telephone conference.
"A new regimen like this could be reducing their (patients') treatment by up to two years or even more," he said, and "promises to be in the order of 90 percent cheaper than the current MDR regimen".
The new drug could reduce from 12,600 to 360 the number of pills an MDR-TB patient would need to take, and eliminate the need for injections.
About nine million people contract tuberculosis every year, and 1.4 million die.
TB is also the lead killer of people with AIDS.
Existing treatments can take anything from six months for "ordinary" TB, to 30 months for those with resistant strains which in some cases are becoming untreatable.
The PaMZ cocktail is in Phase II of clinical trials -- the first stage of human testing.
Early results were published Monday in The Lancet medical journal and presented at the International AIDS Conference in Washington.
The treatment was "well-tolerated and appeared safe", said the study, and seemed to be compatible with AIDS drugs.
The cocktail comprises a candidate TB drug called PA-824, the antibiotic moxifloxacin not yet approved for TB therapy, and an existing TB drug, pyrazinamide -- none of which patients have yet built up resistance to.
Patients can develop drug resistance when they fail to take their medicine as prescribed, but drug-resistant strains can also be directly transmitted from person to person.
PaMZ is being tested in two-month trials with 230 people in South Africa, Tanzania and Brazil -- a precursor to the final Phase III trials with more patients.
The researchers hope to start Phase III next year, but a lack of money may stand in their way.
"We do not have the financial resources to commence that (Phase III) trial," said Spigelman. "We are trying to drum up those resources from various donors."
Andreas Diacon, the trial's principal investigator from the University of Stellenbosch in Cape Town, said that if the testing went according to plan, the drug combination could be commercially available "within four to five years".
The number of MDR cases reported to the World Health Organisation have increased from 29,000 to 53,000 between 2008 and 2010, but the real number could be closer to 300,000, on some estimates.
Mario Raviglione, director of the World Health Organsiations' Stop TB Department, welcomed the findings of the trial which he called a "game-changer".